Beta-trace protein (BTP) is a low–molecular mass protein belonging to the lipocalin protein family, which is more sensitive than serum creatinine for detecting impaired renal function. The aims of the present study were to evaluate whether plasma BTP improves the risk stratification of patients with non–ST-segment elevation acute coronary syndromes and to compare it to cystatin C (CysC), serum creatinine, and estimated glomerular filtration rate. Two hundred twenty-six consecutive patients with non–ST-segment elevation acute coronary syndromes were prospectively included. Blood samples were obtained within 24 hours of hospital admission to measure BTP, CysC, and creatinine. The study end point was all-cause death. Over a median follow-up period of 859 days (interquartile range [IQR] 524 to 1,164), 24 patients (10.6%) died. Decedents had higher concentrations of BTP (1.03 mg/L [IQR 0.89 to 1.43] vs 0.74 mg/L [IQR 0.61 to 0.92], p <0.001), CysC (1.16 mg/L [IQR 0.91 to 1.59] vs 0.90 mg/L [IQR 0.76 to 1.08], p = 0.001), and serum creatinine (1.10 mg/L [IQR 0.87 to 1.46] vs 0.94 mg/L [IQR 0.80 to 1.10], p = 0.004) and a lower mean estimated glomerular filtration rate (60 ± 20 vs 80 ± 24 ml/min/1.73 m 2 , p <0.001). After multivariate adjustment, BTP and CysC were predictors of all-cause death, while estimated glomerular filtration rate and serum creatinine concentrations did not achieve statistical significance. In stratified analyses according to kidney function, elevated BTP and CysC were associated with a higher risk for all-cause death. Reclassification analyses showed that BTP and CysC added complementary information to Global Registry for Acute Coronary Events (GRACE) risk score. In conclusion, BTP and CysC levels were associated with all-cause death risk and modestly improved prognostic discrimination beyond the GRACE risk score in patients with non–ST segment elevation acute coronary syndromes.
Beta-trace protein (BTP) is a low–molecular mass protein belonging to the lipocalin protein family that has been described as a more sensitive marker than serum creatinine in detecting impaired renal function, with comparable performance to cystatin C (CysC). Recently, we described serum BTP level as a powerful predictor of adverse clinical outcomes in the setting of acutely decompensated heart failure, including in patients with estimated glomerular filtration rates (eGFR) in the creatinine-blind zone, and in a similar manner to CysC. In contrast, the prognostic value of BTP levels in acute coronary syndromes (ACS) has not been studied. Therefore, the purposes of the present study were to evaluate the prognostic value of BTP in a population of high-risk patients with non–ST-segment elevation ACS and to compare its prognostic ability to that of CysC and other conventional measures of kidney function, such as creatinine and eGFR. Furthermore, we also wished to evaluate whether BTP or CysC provides additional information to the long-term prognostication scheme of the Global Registry for Acute Coronary Events (GRACE) risk score in patients with non–ST-segment elevation ACS.
Methods
From September 2006 to June 2008, we prospectively enrolled 226 consecutive hospitalized patients with established final diagnoses of high-risk unstable angina or non–ST-segment elevation myocardial infarction. The diagnosis of high-risk non–ST-segment elevation ACS was established on the basis of current guidelines (defined as ischemic symptoms lasting ≥10 minutes and occurring <72 hours before hospital admission and either ST-segment deviation of ≥1 mm or an elevated level of a cardiac biomarker of necrosis). Patients with evidence of hepatic dysfunction or concomitant neoplastic, infectious, connective tissue, or inflammatory diseases were excluded. Furthermore, hospitalization for myocardial infarction, unstable angina, acute decompensated heart failure, or pulmonary embolism in the past 3 months and any cardiac revascularization procedure 1 month before enrollment were also exclusion criteria.
During the entire hospitalization period, baseline clinical characteristics were prospectively recorded, and patients received standard management as recommended for non–ST-segment elevation ACS. Clinical management decisions for each patient were made by the cardiologist responsible, who was unaware of the patient’s BTP and CysC levels. The study was approved by the local ethics committee, and informed consent was obtained from each patient at inclusion.
All blood samples were obtained before coronary angiography within 24 hours of hospital admission (median 8 hours, range 0.5 to 24), and aliquots of serum were immediately stored at −80°C until analyzed. The determination of BTP and CysC was performed using a BN ProSpec analyzer (Dade Behring GmbH, Liederbach, Germany). The intra-assay and interassay coefficients of variation for BTP were 2.8% and 4.7%, respectively. The intra-assay and interassay coefficients of variation for CysC were 2.5% and 2.0%, respectively. Conventional measures of renal function included serum creatinine and eGFR (calculated using the simplified Modification of Diet in Renal Disease equation: 186.3 × [plasma creatinine] −1.154 × age −0.203 × 0.742 [if female]).
After hospital discharge, patients were clinically followed for ≥12 months. The end point of the study was all-cause mortality. Information on deaths was ascertained from available medical records and death certificates. Moreover, all medical records of the patients were carefully reviewed, but when necessary, the patients’ relatives were contacted by telephone to obtain vital status. At the end of the follow-up period, vital status was obtained for all patients.
To compare different predictive values, we constructed areas under the receiver-operating characteristic curves for sensitivity, specificity, positive predictive value, and negative predictive value. We calculated hazard ratios derived from the Cox regression analysis to identify predictors of all-cause death during follow-up. The independent effect of variables on prognosis was calculated using a Cox multivariate regression analysis, incorporating covariates with p values <0.05 in the univariate analysis. To avoid collinearity effects due to their extremely high correlation, serum creatinine and eGFR (r = −0.83, p <0.001), as well as BTP and CysC (r = 0.75, p <0.001), were not entered together in multivariate models. Moreover, given that serum creatinine is one of the parameters included in the GRACE risk score, eGFR and creatinine were not adjusted for this covariate in multivariate analyses. The linearity assumption was tested using Martingale residuals. Log cumulative hazard plots, time-dependent covariates, and Schoenfeld residuals were used to evaluate adherence of the Cox proportional-hazard assumptions. The improvement in predictive accuracy was evaluated by calculating the net reclassification improvement and integrated discrimination improvement, as described by Pencina et al. Finally, the cumulative incidence of all-cause death was estimated according to the Kaplan-Meier method, and the log-rank statistic was used for comparisons. All p values <0.05 were accepted as statistically significant. Statistical analysis was performed using SPSS version 15.0 (SPSS, Inc., Chicago, Illinois) and SAS version 9.2 (SAS Institute Inc., Cary, North Carolina).
Results
The study population consisted of 226 patients with high-risk non–ST-segment elevation ACS. The median plasma BTP concentration for the study group as a whole was 0.77 mg/L (IQR 0.62 to 0.98), the median CysC concentration was 0.91 mg/L (IQR 0.78 to 1.13), the median serum creatinine level was 0.94 mg/dl (IQR 0.81 to 1.11), and the median eGFR was 77 ml/min/1.73 m 2 (IQR 62 to 92). The associations between BTP level and clinical background factors are listed in Table 1 .
| Variable | BTP Level (mg/dl) | p Value | |||
|---|---|---|---|---|---|
| Quartile 1: 0.21–0.62 (n = 58) | Quartile 2: 0.63–0.77 (n = 56) | Quartile 3: 0.78–0.98 (n = 56) | Quartile 4: 0.99–2.33 (n = 56) | ||
| Age (years) | 58 ± 11 | 64 ± 10 | 71 ± 10 | 76 ± 7 | <0.001 |
| Men | 40 (69%) | 31 (55%) | 41 (73%) | 37 (66%) | 0.226 |
| Body mass index (kg/m 2 ) | 30 ± 4 | 29 ± 4 | 28 ± 3 | 29 ± 4 | 0.175 |
| Systolic blood pressure (mm Hg) | 140 (128–161) | 140 (121–160) | 138 (125–164) | 140 (126–166) | 0.934 |
| Heart rate (beats/min) | 79 (70–90) | 78 (66–93) | 77 (68–90) | 78 (62–90) | 0.864 |
| Diabetes mellitus | 24 (41%) | 23 (41%) | 31 (55%) | 28 (50%) | 0.348 |
| Hypertension | 34 (59%) | 45 (80%) | 44 (79%) | 52 (93%) | <0.001 |
| Hyperlipidemia | 26 (45%) | 35 (62%) | 27 (48%) | 31 (55%) | 0.240 |
| Current smoking | 29 (50%) | 15 (27%) | 6 (11%) | 7 (12%) | <0.001 |
| Previous non–ST-segment elevation ACS | 10 (17%) | 18 (32%) | 15 (27%) | 21 (37%) | 0.037 |
| Previous ST-segment elevation myocardial infarction | 5 (9%) | 11 (20%) | 10 (18%) | 14 (25%) | 0.036 |
| Atrial fibrillation/flutter | 1 (2%) | 6 (11%) | 5 (9%) | 7 (12%) | 0.062 |
| Previous stroke | 0 (0%) | 6 (11%) | 5 (9%) | 11 (20%) | 0.001 |
| Previous heart failure | 2 (3%) | 2 (4%) | 1 (2%) | 6 (11%) | 0.121 |
| Ejection fraction (%) | 58 (53–65) | 58 (53–65) | 60 (50–65) | 60 (49–65) | 0.956 |
| Electrocardiographic findings | 0.596 | ||||
| ST-segment depression | 20 (34%) | 18 (32%) | 21 (37%) | 20 (36%) | |
| Transient ST-segment elevation | 8 (14%) | 3 (5%) | 5 (9%) | 3 (5%) | |
| T-wave inversion | 9 (15%) | 11 (20%) | 11 (20%) | 11 (20%) | |
| Laboratory parameters | |||||
| Creatinine (mg/100 ml) | 0.83 (0.75–0.95) | 0.84 (0.76–1.00) | 0.94 (0.87–1.04) | 1.22 (1.09–1.57) | <0.001 |
| eGFR (ml/min/1.73 m 2 ) | 92.1 ± 25.7 | 85.9 ± 19.8 | 77.8 ± 14.2 | 54.8 ± 16.8 | <0.001 |
| Hemoglobin (g/dl) | 14.4 ± 1.9 | 13.9 ± 1.8 | 13.8 ± 1.8 | 13.2 ± 2.0 | 0.013 |
| C-reactive protein (mg/100 ml) | 0.9 (0.4–2.9) | 0.6 (0.3–5.4) | 0.7 (0.2–2.0) | 0.9 (0.4–2.0) | 0.811 |
| Troponin T (ng/ml) | 0.31 (0.01–1.22) | 0.20 (0.01–0.58) | 0.24 (0.04–0.74) | 0.32 (0.05–0.97) | 0.525 |
| GRACE risk score | 115.7 ± 41.6 | 122.5 ± 34.7 | 134.8 ± 30.8 | 148.9 ± 39.9 | <0.001 |
| Number of diseased vessels | <0.001 | ||||
| 1 | 22 (38%) | 20 (36%) | 16 (29%) | 8 (14%) | |
| 2 | 17 (29%) | 11 (20%) | 13 (23%) | 11 (20%) | |
| 3 or left main disease | 7 (12%) | 14 (25%) | 20 (36%) | 23 (41%) | |
| Revascularization | 0.435 | ||||
| Complete | 36 (62%) | 30 (54%) | 31 (55%) | 16 (29%) | |
| Incomplete | 7 (12%) | 11 (20%) | 11 (20%) | 19 (34%) | |
| No | 15 (26%) | 15 (27%) | 14 (25%) | 21 (37%) | |
| Procedure | 0.872 | ||||
| Percutaneous coronary intervention | 38 (65%) | 36 (64%) | 38 (68%) | 31 (55%) | |
| Coronary artery bypass graft | 5 (9%) | 5 (9%) | 4 (7%) | 4 (7%) | |
| Type of stent implanted | 0.088 | ||||
| Drug-eluting stent | 21 (36%) | 26 (46%) | 18 (32%) | 17 (30%) | |
| Bare-metal stent | 17 (29%) | 10 (18%) | 13 (23%) | 7 (12%) | |
| Both | 0 (0%) | 3 (5%) | 7 (12%) | 6 (11%) | |
| Final diagnosis of ACS | 0.744 | ||||
| Unstable angina | 18 (31%) | 19 (34%) | 18 (32%) | 16 (29%) | |
| Non–ST-segment elevation myocardial infarction | 40 (69%) | 37 (66%) | 38 (68%) | 40 (71%) | |
| Treatment at discharge | |||||
| Aspirin | 56 (97%) | 54 (96%) | 54 (96%) | 48 (96%) | 0.166 |
| Clopidogrel | 45 (78%) | 45 (80%) | 48 (86%) | 44 (79%) | 0.290 |
| β blockers | 50 (86%) | 50 (89%) | 49 (87%) | 42 (75%) | 0.321 |
| Angiotensin-converting enzyme inhibitors/angiotensin receptor blockers | 51 (88%) | 45 (80%) | 51 (91%) | 45 (80%) | 0.937 |
| Statins | 54 (93%) | 54 (96%) | 54 (96%) | 51 (91%) | 0.264 |
| Acenocoumarol | 3 (5%) | 3 (5%) | 5 (9%) | 4 (7%) | 0.465 |
Plasma BTP and CysC concentrations were strongly correlated with each other (r = 0.75, p <0.001) and moderately with serum creatinine (r = 0.57, p <0.001 for both), eGFR (r = −0.59 for BTP and r = −0.68 for CysC, p <0.001 for both), age (r = 0.60 for BTP and r = 0.53 for CysC, p <0.001 for both), and GRACE risk score (r = 0.35 for BTP and r = 0.43 for CysC, p <0.001 for both). Plasma BTP and CysC concentrations were also weakly correlated with hemoglobin (r = −0.23 for BTP and r = −0.28 for CysC, p <0.001 for both). In multiple linear regression analyses, eGFR and age were the main independent predictors of BTP (β = 0.45 and β = 0.33, respectively, p <0.001 for both) and CysC (β = 0.54 and β = 0.58, respectively, p <0.001 for both). Current smoking status was associated with increased BTP and CysC concentrations in univariate analyses, but after multivariate adjustment, it was an independent predictor only of CysC levels (β = 0.16, p = 0.004). The fit (adjusted R 2 ) of models was 0.48 (p <0.001) for log 10 BTP and 0.52 (p <0.001) for log 10 CysC.
Over the study period (median 859 days, IQR 524 to 1.164), a total of 24 patients (10.6%) died. Compared to survivors, decedents had higher concentrations of BTP (1.03 mg/L [IQR 0.89 to 1.43] vs 0.74 mg/L [IQR 0.61 to 0.92], p <0.001), CysC (1.16 mg/L [IQR 0.91 to 1.59] vs 0.90 mg/L [IQR 0.76 to 1.08], p = 0.001), and serum creatinine (1.10 mg/L [IQR 0.87 to 1.46] vs 0.94 mg/L [IQR 0.80 to 1.10], p = 0.004) but a lower mean eGFR (60 ± 20 vs 80 ± 24 ml/min/1.73 m 2 , p <0.001).
BTP and CysC concentrations had similar ability to discriminate between decedents and survivors, while serum creatinine levels had overall lower prognostic accuracy ( Table 2 ). In univariate Cox regression analysis, hemoglobin, previous non–ST-segment elevation ACS, and GRACE risk score, as well as all measures of kidney function, were significantly associated with a higher risk for all-cause death. However, after adjusting for these covariates in multivariate Cox regression analyses, only GRACE risk score, BTP, and CysC remained significant predictors of all-cause death ( Table 3 ). As detailed in Figure 1 , this remained unchanged when these markers were evaluated as quartiles. Furthermore, constraining the model to 1 year of follow up, BTP and CysC showed a similar prognostic role: BTP (per 1 mg/dl, hazard ratio 6.12, 95% confidence interval [CI] 1.44 to 25.9, p = 0.014) and CysC (per 1 mg/dl, hazard ratio 4.08, 95% CI 1.09 to 16.9, p = 0.042). In contrast, neither serum creatinine nor eGFR was independently associated with 1-year all-cause death (p >0.11).
| Variable | Area Under the Curve | 95% CI | Cutoff | Sensitivity | Specificity | Positive Predictive Value | Negative Predictive Value | p Value ⁎ |
|---|---|---|---|---|---|---|---|---|
| BTP | 0.79 | 0.73–0.84 | 0.88 | 0.83 | 0.71 | 0.25 | 0.97 | |
| CysC | 0.75 | 0.69–0.80 | 1.13 | 0.54 | 0.81 | 0.26 | 0.94 | 0.08 |
| Creatinine | 0.68 | 0.62–0.74 | 1.03 | 0.63 | 0.70 | 0.20 | 0.94 | 0.03 |
| eGFR | 0.74 | 0.67–0.79 | 69 | 0.75 | 0.71 | 0.22 | 0.96 | 0.05 |
| GRACE risk score | 0.73 | 0.67–0.79 | 139 | 0.71 | 0.68 | 0.22 | 0.95 | 0.05 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
