Clinica Chimica Acta 444 (2015) 86–91
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
A head-to-head comparison of homocysteine and cystatin C as pre-procedure predictors for contrast-induced nephropathy in patients undergoing coronary computed tomography angiography Suhua Li a, Xixiang Tang b, Long Peng a, Yanting Luo a, Yunyue Zhao a, Lin Chen a, Ruimin Dong a, Jieming Zhu a, Yanming Chen b, Jinlai Liu b,⁎ a b
Department of Cardiology, The Third Affiliated Hospital, Sun Yat-sen University, Tian-he Road, Guangzhou 510630, China Advanced Medical Center, The Third Affiliated Hospital, Sun Yat-sen University, Tian-he Road, Guangzhou 510630, China
a r t i c l e
i n f o
Article history: Received 22 November 2014 Received in revised form 5 February 2015 Accepted 5 February 2015 Available online 14 February 2015 Keywords: Contrast-induced nephropathy Computed tomography angiography Homocysteine Cystatin C Creatinine
a b s t r a c t Background: Homocysteine is a potential predictor for contrast-induced nephropathy (CIN). We aimed to compare homocysteine with cystatin C as pre-procedure predictors for CIN in patients undergoing coronary computed tomography angiography (CCTA). Methods: A total of 580 consecutive patients were enrolled. Concentrations of plasma homocysteine and serum cystatin C were measured before CCTA. CIN is defined as an elevation of creatinine by ≥ 25% or ≥ 0.5 mg/dl from baseline within 48 h. Receiver operating characteristic curves, Pearson correlation coefficients and logistic regression analysis were used to evaluate the efficiency of potential predictors. Results: Fifty-seven (9.83%) patients developed CIN. Concentrations of homocysteine (19.35 ± 4.32 μmol/l vs. 13.42 ± 3.96 μmol/l, p b 0.001) and cystatin C (1.20 ± 0.21 mg/dl vs. 0.99 ± 0.15 mg/dl, p b 0.001) increased significantly in CIN subjects. CIN was predicted by homocysteine (AUC 0.829, p b 0.001) and cystatin C (AUC 0.774, p b 0.001), while creatinine was not predictive. Both homocysteine and cystatin C had positive correlation with ΔCreatinine48h-0 (p b 0.001) and negative correlation with ΔeGFR48h-0 (p b 0.001). Regression analysis confirmed that increased baseline homocysteine [OR: 1.262 (1.123, 2.554), p b 0.001] and cystatin C [OR: 1.565 (1.380, 1.775), p b 0.001] were independent predictors for CIN. Conclusions: Homocysteine, with similar predictive value compared to cystatin C, was an independent biomarker for predicting CIN before CCTA examination. © 2015 Elsevier B.V. All rights reserved.
1. Introduction With the rapid advances in mechanical devices and interventional cardiology, a growing number of patients with suspected or known coronary artery disease are undergoing coronary computed tomography angiography (CCTA), conventional angiography and percutaneous coronary intervention in clinical practice. The extensive administration of contrast agents during these procedures has become a highly focused issue globally [1–3]. It is reported that contrast-induced nephropathy (CIN), a serious complication associated with contrast media administration, is the third most common cause of hospital-acquired acute kidney injury and accounts for about 11–12% of cases [4]. CIN is considered to be connected with increased morbidity, prolonged
⁎ Corresponding author at: Department of Cardiology, the Third Affiliated Hospital, Sun Yat-sen University, Tian-he Road, Guangzhou 510630, China. Tel.: +86 20 85252168; fax: +86 20 85252874. E-mail address: [email protected] (J. Liu).
http://dx.doi.org/10.1016/j.cca.2015.02.019 0009-8981/© 2015 Elsevier B.V. All rights reserved.
hospitalization, increased risk of complications, potential need for dialysis and increased mortality rate [4]. Conventionally, CIN is defined as an elevation of serum creatinine (sCr) of N 25% or ≥ 0.5 mg/dl (44 μmol/l) from baseline within 48 h after a diagnostic or interventional procedure in the absence of an alternative etiology [5–7]. However, sCr is not an ideal indicator for CIN, since it is secreted by renal tubules and influenced by muscle mass [8]. Therefore, more sensitive markers of CIN are desired. Recently, cystatin C, a cysteine protease inhibitor which has a constant production rate irrespective of muscle mass and a plasma concentration determined by glomerular filtration alone, has been well demonstrated to be superior to sCr for predicting CIN in a number of studies [4,9–13]. On the other hand, hyperhomocysteinemia is considered to be independently associated with a greater risk of CIN after angiography or angioplasty in two previous studies [14,15]. An explanation is that hyperhomocysteinemia induces oxidative stress, cellular apoptosis and endothelial dysfunction, which shares the proposed pathophysiologic mechanisms of CIN. However, the predicting value of plasma homocysteine comparing to serum cystatin C (an ideal predictor of CIN) has never been evaluated.
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
2. Materials and methods 2.1. Study population The current study was approved by the Institutional Review Board of the hospital. The written informed consents were obtained. From January, 2012 to July, 2014, consecutive inpatients receiving 320-slice CCTA (Aquilion ONE; Toshiba Medical Systems) for any reason were enrolled. The indications for performing CCTA were 1) evaluation of suspected coronary artery disease (depending on angina-like symptoms, cardiogenic syncope, elevated cardiovascular risk, abnormal echocardiogram, positive stress ECG test, positive myocardial perfusion scintigraphies), 2) determination of the patency of bypass grafts or stents, 3) evaluation of cardiomyopathy, and 4) preoperative evaluation for non-coronary surgery. Patients were excluded if they had any of the following conditions: 1) impaired renal function (eGFR b60 ml/min per 1.73 m2), 2) exposure to nephrotoxic drugs prior to or during the study period, 3) allergy to iodine-containing contrast medium, 4) contraindication for β-blockers, and 5) pregnancy. 2.2. Serum measurements Blood samples for evaluations of plasma homocysteine and serum cystatin C were collected at baseline, while samples for estimations of sCr were collected at 0, 24 and 48 h of CCTA. Blood samples were immediately sent to the central laboratory of the hospital, and analyzed on the Siemens ADVIA Centaur for measurement of homocysteine, or on the Hitachi 7180 clinical chemistry analyzer for measurement of cystatin C and sCr. Plasma homocysteine was measured using a competitive immunoassay with direct chemiluminescence detection. Serum cystatin C was measured via an immunoturbidimetric technique. sCr was estimated by Jaffe's method. Estimated glomerular filtration rate (eGFR) was calculated by the Modification of Diet in Renal Disease (MDRD) study equation [16]. Baseline concentrations of blood urea nitrogen (BUN), total cholesterol, triglyceride, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), uric acid, albumin and fasting glucose were also measured using ITACHI 7180 chemistry analyzer. Baseline hemoglobin was measured by the cyanmethemoglobin method. According to the definition of CIN relied on the changes of sCr within 48 h after CCTA, patients were assigned to either CIN(+) or CIN(−) group. 2.3. Coronary computed tomography angiography Before CCTA examination, 500 ml oral fluid was given to each patient. After CCTA, patients took unrestrictively oral fluid as much as possible. Oral metoprolol tablets (50–100 mg, Astrazeneca) were given to patients with a resting heart rate N 70 bpm. Isovue-370 (50–100 ml, Bracco Diagnostics) was injected intravenously at a flow rate of 6.0 ml s−1, followed by 20 ml saline flush at a flow rate of 4.0 ml s− 1. 320-slice CCTA was performed with 0.5-mm detector element, 350 ms of gantry rotation time, and up to 16 cm of coverage in Z direction. Tube voltages were set at 100–135 kV and the maximal tube currents were set at 400–580 mA. 2.4. Statistical analysis Statistical analysis was performed by using SPSS 18.0. Categorical variables were expressed as percentages, and continuous variables were presented as means ± SD. Differences of continuous variables between groups were evaluated by Student's t test, while differences of categorical variables were evaluated by Pearson chi-square test. A p b 0.05 was considered to be statistically significant, unless otherwise specified. Receiver operating characteristic (ROC) curves were drawn to detect the optimum cut-offs of parameters and their sensitivity and specificity in predicting CIN. The optimum cut-off is defined as the
87
point where one gets a maximum value of Youden's index (J = sensitivity + specificity − 1) in the ROC curve. Pearson correlations were used to explore the associations between potential biomarkers and the changes of sCr and eGFR after CCTA. In addition, a univariate logistic regression analysis was performed through potential predictive factors. And then, a multivariate logistic regression analysis was used to assess the correlation among the parameters which got a level of p b 0.05 in the univariate analysis.
3. Results 3.1. Patient characteristics A total of 580 eligible patients were enrolled. Fifty-seven (9.83%) of them developed CIN after the infusion of contrast medium, while 523 (90.17%) of them did not. Among patients developing CIN, 70.18% of them were males, which was higher than the percentage in the CIN(−) group. CIN(+) patients were significantly older than CIN(−) ones on average (67.2 ± 9.4 vs. 62.6 ± 10.9 y, p = 0.002). In addition, CIN(+) patients had larger portion of diabetes (71.9% vs. 57.9%, p = 0.041), higher total cholesterol level (139.0 ± 18.1 mg/dl vs. 132.4 ± 19.6 mg/dl, p = 0.016), and poor HbA1c level (6.84 ± 1.10% vs. 6.53 ± 1.04%, p = 0.031) than those of CIN(−) patients. However, the other baseline profiles showed no significant difference between CIN(+) and CIN(−) groups (Table 1).
3.2. Pre-CCTA concentrations of biomarkers At baseline (Fig. 1), the levels of sCr (0.94 ± 0.06 vs. 0.93 ± 0.09 mg/dl, p = 0.556), eGFR (80.86 ± 6.86 1.73 m2 vs. 83.35 ± 10.25 ml/min · 1.73 m2 , p = 0.085) and BUN (15.58 ± 2.61 vs. 14.97 ± 2.40 mg/dl, p = 0.074) were similar between CIN(+) and CIN(−) groups, whereas pre-CCTA concentrations of homocysteine (19.35 ± 4.32 vs. 13.42 ± 3.96 μmol/l, p b 0.001) and cystatin C (1.20 ± 0.21 vs. 0.99 ± 0.15 mg/dl, p b 0.001) in CIN(+) patients were significantly higher than those in CIN(−) subjects.
Table 1 Characteristics of eligible patients. Characteristics
CIN(+) (n = 57)
CIN(−) (n = 523)
p
Male/Female, n1/n2 Age, y Body mass index, kg/m2 Hypertension, n Diabetes, n Smoker, n Alcohol, n History of myocardial infarction, n Single-/double-/triple-vessel disease, n1/n2/n3 Contrast, ml Fluid intake after CCTA, ml Statin, n Hemoglobin, g/l Total cholesterol, mg/dl Triglyceride, mg/dl HDL-C, mg/dl LDL-C, mg/dl Albumin, g/l Uric acid, μmol/l Fasting glucose, mmol/l HbA1c, % Left ventricular ejection fraction, %
40/17 67.2 ± 9.4 23.01 ± 3.01 18 41 39 43 4
288/235 62.6 ± 10.9 23.44 ± 2.66 152 303 355 333 25
0.029a 0.002b NSb NSa 0.041a NSa NSa NSa
12/9/6
89/80/53
0NSa
74.3 ± 9.7 1240 ± 184 46 118.7 ± 12.8 139.0 ± 18.1 100.0 ± 14.6 31.5 ± 3.5 78.9 ± 8.4 36.4 ± 3.1 332.6 ± 64.7 7.1 ± 1.4 6.84 ± 1.10 59.5 ± 8.1
73.5 ± 8.1 1229 ± 200 458 120.2 ± 15.4 132.4 ± 19.6 102.2 ± 13.6 30.8 ± 3.49 77.5 ± 9.8 36.3 ± 3.4 347.1 ± 73.8 6.7 ± 1.5 6.53 ± 1.04 59.0 ± 7.7
NSb NSb NSa NSb 0.016b NSb NSb NSb NSb NSb NSb 0.031b NSb
a b
p value calculated by Pearson χ2 test. p value calculated by Student's t test.
88
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
Fig. 1. Baseline concentrations of homocysteine (a), cystatin C (b), sCr (c), eGFR (d) and BUN (e). Homocysteine and cystatin C were significantly higher in the CIN(+) group than the CIN(−) group, whereas sCr, eGFR and BUN had no differences.
3.3. ROC curves of potential biomarkers ROC curves of potential biomarkers for predicting CIN were drawn in Fig. 2. The areas under the curve (AUC) of homocysteine (AUC 0.829, 95% CI 0.771–0.886) and cystatin C (AUC 0.774, 95% CI 0.700–0.848) were significantly higher than those of sCr (AUC 0.569, 95% CI 0.502– 0.636), eGFR (AUC 0.571, 95% CI 0.486–0.656) and BUN (AUC 0.571, 95% CI 0.487–0.655). Thus, both pre-CCTA homocysteine (p b 0.001)
and cystatin C (p b 0.001) had predictive values for CIN, while sCr (p = 0.087), eGFR (p = 0.078) and BUN (p = 0.079) had no predictive values. At a optimum cut-off of 16.65 μmol/l, plasma homocysteine had a maximum sensitivity of 71.9% and specificity of 75.0% for predicting CIN. At an optimum cut-off of 1.05 mg/dl, serum cystatin C had a maximum sensitivity of 68.4% and specificity of 71.7% for predicting CIN. 3.4. Associations between potential biomarkers and the changes of sCr and eGFR As shown in Table 2, pre-CCTA plasma homocysteine had positive correlation with ΔsCr48h-0 (r = 0.356, p b 0.001) and negative correlation with ΔeGFR48h-0 (r = −0.307, p b 0.001). However, homocysteine had no correlations with ΔsCr24h-0 and ΔeGFR24h-0. Unlike that, preCCTA serum cystatin C had positive correlations with both ΔsCr24h-0 (r = 0.164, p b 0.001) and ΔsCr48h-0 (r = 0.405, p b 0.001), and negative correlations with both ΔeGFR24h-0 (r = − 0.163, p b 0.001) and ΔeGFR48h-0 (r = − 0.410, p b 0.001). Although baseline sCr, eGFR and BUN also had some correlations with the changes of sCr and eGFR at either 24 h or 48 h after CCTA, the correlation coefficients were relatively low (r = −0.197 to 0.185). 3.5. Logistic regression analysis
Fig. 2. The receiver operating characteristic (ROC) curves of baseline homocysteine, cystatin C, sCr, eGFR and BUN. Both homocysteine (AUC 0.829, p b 0.001; optimum cutoff 16.65 μmol/l, sensitivity 71.9%, specificity 75.0%) and cystatin C (AUC 0.774, p b 0.001; optimum cut-off 1.05 mg/dl, sensitivity 68.4%, specificity 71.7%), had predictive values for CIN, while sCr (AUC 0.569, p = 0.087), eGFR (AUC 0.571, p = 0.078) and BUN (AUC 0.571, p = 0.079) had none.
In the univariate logistic regression analysis, gender, age, total cholesterol, sCr, eGFR, homocysteine and cystatin C were found to be potentially correlated with the development of CIN (p b 0.05) (Table 3). As shown in Table 4, the multivariate logistic regression analysis confirmed that increased baseline homocysteine (odds ratio 1.262, 95% CI 1.123– 2.554, p b 0.001) and cystatin C (odds ratio 1.565, 95% CI 1.380–1.775, p b 0.001) were independent predictors for the development of CIN, while baseline sCr (odds ratio 0.987, 95% CI 0.280–1.989, p = 0.195) and eGFR (odds ratio 0.966, 95% CI 0.503–2.162, p = 0.251) were not independent predictors. In addition, males (odds ratio 1.363, 95% CI
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
89
Table 2 Pearson correlations. ΔsCr48h-0
ΔsCr24h-0
ΔeGFR24h-0
ΔeGFR48h-0
Biomarkers
Correlation
p
Correlation
p
Correlation
p
Correlation
p
Homocysteine Cystatin C sCr eGFR BUN
0.101 0.164 0.056 0.064 0.057
0.015a b0.001a NS NS NS
0.356 0.405 0.185 −0.041 0.106
b0.001a b0.001a b0.001a NS 0.011a
−0.079 −0.163 0.124 −0.103 −0.065
NS b0.001a 0.003a 0.021a NS
−0.307 −0.410 0.025 −0.197 −0.153
b0.001a b0.001a NS b0.001a b0.001a
a
p ≤ 0.05, indicating significant correlations between biomarkers and the changes of sCr or eGFR.
1.022–1.364, p = 0.005), elders (odds ratio 1.197, 95% CI 1.047–1.150, p b 0.001), and patients with higher total cholesterol levels (odds ratio 1.026, 95% CI 1.004–1.049, p = 0.023) seemed to have higher risks for developing CIN. 4. Discussion The present prospective study was the first “head-to-head” trial to compare plasma homocysteine with serum cystatin C for the prediction of CIN in patients undergoing CCTA. The authors found that both plasma homocysteine and serum cystatin C were independent biomarkers for pre-CCTA prediction of CIN. At respective optimum cut-offs (16.65 μmol/l for homocysteine, and 1.05 mg/dl for cystatin C), plasma homocysteine and serum cystatin C had similar sensitivity (71.9% vs. 68.4%) and specificity (75.0% vs. 71.7%) for predicting CIN. The occurrence of CIN in the present study was 9.83%, which is similar to those of previous trials [5,9,10]. Those patients with positive results on CCTA (narrowing of coronary lumen ≥50%) or high clinical probability of coronary artery disease tended to undergo further angiography and percutaneous coronary intervention in clinical routine, leading to the overexposure of contrast media, deterioration of renal function and adverse outcomes. Although several nonpharmacological and pharmacological approaches have been evaluated for the management of CIN, there is no definitive treatment available for CIN, making
Table 3 Univariate logistic regression analysis of potential factors determining CIN. Factors
Male Age Body mass index Hypertension Diabetes Smoker Alcohol Coronary artery disease Contrast volume Statin Hemoglobin Total cholesterol Triglyceride HDL-C LDL-C Albumin Fasting glucose Uric acid HbA1c Left ventricular ejection fraction Fluid intake after CCTA BUN Serum creatinine eGFR Cystatin C Homocysteine
Odds ratio
1.575 1.096 0.991 1.764 2.485 1.685 1.586 1.127 1.012 0.989 1.012 1.029 0.997 1.031 1.042 0.965 1.298 0.998 1.457 0.993 0.999 0.900 1.178 1.124 1.885 1.591
95% CI
p
Lower
Upper
1.013 1.037 0.797 0.629 0.934 0.535 0.542 0.450 0.959 0.746 0.981 1.003 0.962 0.909 0.995 0.850 0.942 0.991 0.954 0.940 0.997 0.740 1.076 1.035 1.048 1.387
2.394 1.158 1.232 4.950 6.613 5.305 4.645 2.820 1.086 1.355 1.043 1.055 1.034 1.170 1.091 1.097 1.788 1.004 2.226 1.048 1.002 1.096 2.048 1.222 4.173 1.826
0.049a 0.001a NS NS NS NS NS NS NS NS NS 0.028a NS NS NS NS NS NS NS NS NS NS 0.014a 0.006a b0.001a b0.001a
a Potential factors on the univariate analysis were added to multivariate logistic regression analysis for assessing the correlation among the parameters.
prevention the corner stone of management [17]. With the main goals of helping clinicians weight the risk of the exposure versus its benefit, risk scoring and stratification is a matter of concern for CIN prevention before contrast media administration. Recently, several risk scoring and stratification systems for CIN have been developed, such as the Bartholomew's score system, the CR4EATME3AD3 score, and the Mehran's score system [18–21]. These models were usually established by the conventional risk factors for CIN, including elderly age, impaired renal function, urgent coronary procedure, use of IABP, large contrast volume, left ventricular dysfunction, history of myocardial infarction, diabetes mellitus, hypertension, hypotension, peripheral vascular disease, anemia, hypoalbuminemia, dehydration, and cirrhosis. The present study confirmed that males, elders, and patients with higher total cholesterol levels seemed to have higher risks for the development of CIN. However, models based on conventional risk factors only were not sensitive and reliable enough to predictive CIN. One typical example is that CIN might develop even in subjects with normal sCr at baseline, which is an intrinsic flaw of sCr so-called “creatinine blind range” [22]. There is usually a 24–48 h delay between contrast exposure and the change in sCr. This delay makes sCr a late and unreliable indicator for the acute change of renal function [8]. Therefore, it is important to find more sensitive and accurate markers to predict CIN before the procedure. Fortunately, several biomarkers have been under evaluation, such as neutrophil gelatinase-associated lipocalin, cystatin C, interleukin-18, liver-type fatty acid-binding protein, and kidney injury molecule-1 [9,17,23,24]. Among them, cystatin C receives most interest in the searches for ideal predictors of CIN. In fact, cystatin C has been demonstrated to be an independent predictor for CIN in the present and previous studies [4,9–13]. Investigators found that CIN was predicted by baseline cystatin C independently, whereas baseline sCr were not predictive. There are several theoretical reasons why cystatin C is superior to sCr for predicting CIN [5,8,12,13,25]. sCr is a byproduct of muscle breakdown. It might cause the overestimation of eGFR in individuals with lower sCr generation due to loss of muscle mass. In addition, sCr can be actively secreted by renal tubules, which further reduces the value of sCr to estimate eGFR. Unlike that, cystatin C has a constant production rate irrespective of muscle mass, and a plasma concentration determined by glomerular filtration alone, which is not subject to tubular secretion.
Table 4 Multivariate logistic regression analysis of potential factors determining CIN. Factors
Male Age Total cholesterol sCr eGFR Cystatin C Homocysteine a
Odds ratio
1.363 1.097 1.026 0.987 0.966 1.565 1.262
95% CI
p
Lower
Upper
1.022 1.047 1.004 0.280 0.503 1.380 1.123
4.364 1.150 1.049 1.989 2.162 1.775 2.554
0.005a b0.001a 0.023a NS NS b0.001a b0.001a
p ≤ 0.05, indicating independent predictive factors for the development of CIN.
90
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
On the contrary, plasma homocysteine receives less attention among the studies of CIN. Up to now, only 2 studies have investigated the relationship between homocysteine and the risk of CIN [14,15]. In a cohort of 572 patients who underwent percutaneous coronary intervention, Kim et al. found that the incidence of CIN was significantly greater in patients in the third homocysteine tertile (from lowest to highest, 4.7%, 7.3%, and 24.2%, p b 0.001). Hyperhomocysteinemia was an independent risk factor for CIN (odds ratio 1.70, 95% CI 1.07–2.71, p = 0.025) [15]. Similarly, Barbieri et al. investigated 876 patients with impaired renal function (eGFR b60 ml/min) who underwent coronary angiography and/or angioplasty, and found a significant relationship between homocysteine and CIN (odds ratio 1.68, 95% CI 1.09–2.59, p = 0.019) [14]. Therefore, homocysteine is a new potential biomarker for CIN, which deserves more attention in the near future. However, the predicting value of plasma homocysteine comparing to serum cystatin C, the ideal predictor of CIN, has never been evaluated. As the first attempt, the present study confirmed that plasma homocysteine, with similar sensitivity and specificity at the optimum cut-off compared to cystatin C, was an independent biomarker for the prediction of CIN after correction of the baseline confounding factors. Unlike Kim's or Barbier's studies, the current trial focused on patients receiving CCTA, a widespread non-invasive approach for evaluating coronary artery trees. It is expected to help a large number of patients weight the risk of CCTA versus its benefit. The link between homocysteine and the risk of CIN can be partly explained by the intrinsic pathophysiologic mechanisms they shared. Although the definite mechanism of CIN is not well-established, several mechanisms have been proposed: 1) direct toxicity of contrast media which is related to free radicals and oxidative stress; 2) apoptosis of renal cells induced by contrast media; and 3) renal medullary hypoxia [17,26,27]. Coincidentally, homocysteine is a potent excitatory neurotransmitter that binds to the N-methyl-D-aspartate receptor, leading to oxidative stress, cellular apoptosis, and endothelial dysfunction. Therefore, it is reasonable to hypothesize that hyperhomocysteinemia and CIN might have a causal relationship. However, the present observational study is unable to give a definite answer. N-acetylcysteine, an antioxidant which lowers homocysteine, has been shown to protect patients from CIN after coronary angiographic procedures [28,29]. The protective role of N-acetylcysteine is attributed to the inhibition of oxygen free radical production upon contrast media exposure. Unfortunately, none of the previous studies were primarily designed to lower homocysteine, and the change of homocysteine after N-acetylcysteine treatment had not been well evaluated. Therefore, the relationship between homocysteine-lowering therapy and CIN remains uncertain. Further prospective interventional study is needed in the near future to clarify whether homocysteine-lowering therapy among patients with hyperhomocysteinemia reduces the incidence of CIN or not. 4.1. Clinical implications The present study provides available data for the prediction of CIN in patients receiving CCTA, and is a good supplement to the previous risk score models. Based on the levels of serum cystatin C and plasma homocysteine, physicians can obtain more information for the prevention of CIN prior to CCTA, by using less contrast agent, sufficient hydration, and potential preventive agents. 4.2. Limitations It is a single center study, and trials on more mixed populations are warranted. The present study did not include patients with eGFR b60 ml/min per 1.73 m2, since these patients were not recommended to receive CCTA. Thus, the predictive value of sCr might be underestimated in the present study, which might limit its application in clinical practice. Furthermore, the authors only examined the changes
of biomarkers for 48 h, and long-term predictive values of biomarkers should be investigated further. 4.3. Conclusions CIN is a common potential complication after contrast administration in CCTA. Plasma homocysteine, with similar value compared to serum cystatin C, was an independent biomarker for pre-CCTA prediction of CIN. Studies on homocysteine-lowering therapy are warranted for reducing the incidence of CIN.
Abbreviations CIN sCr CCTA LVEF
contrast-induced nephropathy serum creatinine coronary computed tomography angiography left ventricular ejection fraction
References [1] EI-Hajjar M, Bashir I, Khan M, Min J, Torosoff M, DeLago A. Incidence of contrastinduced nephropathy in patients with chronic renal insufficiency undergoing multidetector computed tomographic angiography treated with preventive measures. Am J Cardiol 2008;102:353–6. [2] Ko B, Garcia S, Mithani S, Tholakanahalli V, Adabag S. Risk of acute kidney injury in patients who undergo coronary angiography and cardiac surgery in close succession. Eur Heart J 2012;33:2065–70. [3] Leoncini M, Toso A, Maioli M, Tropeano F, Villani S, Bellandi F. Early high-dose rosuvastatin for contrast-induced nephropathy prevention in acute coronary syndrome: results from the PRATO-ACS Study (Protective Effect of Rosuvastatin and Antiplatelet Therapy On contrast-induced acute kidney injury and myocardial damage in patients with Acute Coronary Syndrome). J Am Coll Cardiol 2014;63: 71–9. [4] Maliborski A, Zukowski P, Nowicki G, Bogusławska R. Contrast-induced nephropathy—a review of current literature and guidelines. Med Sci Monit 2011; 17:RA199–204. [5] Ribichini F, Gambaro G, Graziani MS, et al. Comparison of serum creatinine and cystatin C for early diagnosis of contrast-induced nephropathy after coronary angiography and interventions. Clin Chem 2012;58:458–64. [6] Aspelin P, Aubry P, Fransson SG, Strasser R, Willenbrock R, Berg KJ. Nephrotoxic effects in high-risk patients undergoing angiography. N Engl J Med 2003;348: 491–9. [7] Mohammed NM, Mahfouz A, Achkar K, Rafie IM, Hajar R. Contrast-induced nephropathy. Heart Views 2013;14:106–16. [8] Dalton RN. Serum creatinine and glomerular filtration rate: perception and reality. Clin Chem 2010;56:687–9. [9] Padhy M, Kaushik S, Girish MP, Mohapatra S, Shah S, Koner BC. Serum neutrophil gelatinase associated lipocalin (NGAL) and cystatin C as early predictors of contrast-induced acute kidney injury in patients undergoing percutaneous coronary intervention. Clin Chim Acta 2014;435:48–52. [10] Alharazy SM, Kong N, Saidin R, et al. Serum neutrophil gelatinase-associated lipocalin and cystatin C are early biomarkers of contrast-induced nephropathy after coronary angiography in patients with chronic kidney disease. Angiology 2014;65:436–42. [11] Wacker-Gußmann A, Bühren K, Schultheiss C, et al. Prediction of contrast-induced nephropathy in patients with serum creatinine levels in the upper normal range by cystatin C: a prospective study in 374 patients. AJR Am J Roentgenol 2014;202: 452–8. [12] Nozue T, Michishita I, Mizuguchi I. Predictive value of serum cystatin C, β2microglobulin, and urinary liver-type fatty acid-binding protein on the development of contrast-induced nephropathy. Cardiovasc Interv Ther 2010;25:85–90. [13] Briguori C, Visconti G, Rivera NV, et al. Cystatin C and contrast-induced acute kidney injury. Circulation 2010;121:2117–22. [14] Barbieri L, Verdoia M, Schaffer A, et al. Elevated homocysteine and the risk of contrastinduced nephropathy: a cohort study. Angiology 2014 [pii: 0003319714533401, Epub ahead of print]. [15] Kim SJ, Choi D, Ko YG, et al. Relation of homocysteinemia to contrast-induced nephropathy in patients undergoing percutaneous coronary intervention. Am J Cardiol 2011;108:1086–91. [16] Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of diet in renal disease study group. Ann Intern Med 1999; 130:461–70. [17] Mohammed NM, Mahfouz A, Achkar K, Rafie IM, Hajar R. Contrast-induced Nephropathy. Heart Views 2013;14:106–16. [18] Bartholomew BA, Harjai KJ, Dukkipati S, et al. Impact of nephropathy after percutaneous coronary intervention and a method for risk stratification. Am J Cardiol 2004; 93:1515–9.
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91 [19] Fu N, Li X, Yang S, et al. Risk score for the prediction of contrast induced nephropathy in elderly patients undergoing percutaneous coronary intervention. Angiology 2013; 64:188–94. [20] Mehran R, Aymong ED, Nikolsky E, et al. A simple risk score for prediction of contrast-induced nephropathy after percutaneous coronary intervention: development and initial validation. J Am Coll Cardiol 2004;44:1393–9. [21] Tziakas D, Chalikias G, Stakos D, et al. Development of an easily applicable risk score model for contrast-induced nephropathy prediction after percutaneous coronary intervention: a novel approach tailored to current practice. Int J Cardiol 2013;163: 46–55. [22] Takeuchi T, Isobe S, Sato K, et al. Cystatin C: a possible sensitive marker for detecting potential kidney injury after computed tomography coronary angiography. J Comput Assist Tomogr 2011;35:240–5. [23] Ling W, Zhaohui N, Ben H, et al. Urinary IL-18 and NGAL as early predictive biomarkers in contrast-induced nephropathy after coronary angiography. Nephron Clin Pract 2008;108:c176–81.
91
[24] Nogare AL, Dalpiaz T, Veronese FJ, Gonçalves LF, Manfro RC. Noninvasive analyses of kidney injury molecule-1 messenger RNA in kidney transplant recipients with graft dysfunction. Transplant Proc 2012;44:2297–9. [25] Hatano M. Clinical study of tubular creatinine secretion in renal dysfunction. Jpn J Nephrol 1991;33:1097–104. [26] Romano G, Briguori C, Quintavalle C, et al. Contrast agents and renal cell apoptosis. Eur Heart J 2008;29:2569–76. [27] Hizoh I, Haller C. Radiocontrast-induced renal tubular cell apoptosis: hypertonic versus oxidative stress. Invest Radiol 2002;37:428–34. [28] Kay J, Chow WH, Chan TM, et al. Acetylcysteine for prevention of acute deterioration of renal function following elective coronary angiography and intervention: a randomized controlled trial. JAMA 2003;289:553–8. [29] Marenzi G, Assanelli E, Marana I, et al. N-acetylcysteine and contrast-induced nephropathy in primary angioplasty. N Engl J Med 2006;354:2773–82.
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
A head-to-head comparison of homocysteine and cystatin C as pre-procedure predictors for contrast-induced nephropathy in patients undergoing coronary computed tomography angiography Suhua Li a, Xixiang Tang b, Long Peng a, Yanting Luo a, Yunyue Zhao a, Lin Chen a, Ruimin Dong a, Jieming Zhu a, Yanming Chen b, Jinlai Liu b,⁎ a b
Department of Cardiology, The Third Affiliated Hospital, Sun Yat-sen University, Tian-he Road, Guangzhou 510630, China Advanced Medical Center, The Third Affiliated Hospital, Sun Yat-sen University, Tian-he Road, Guangzhou 510630, China
a r t i c l e
i n f o
Article history: Received 22 November 2014 Received in revised form 5 February 2015 Accepted 5 February 2015 Available online 14 February 2015 Keywords: Contrast-induced nephropathy Computed tomography angiography Homocysteine Cystatin C Creatinine
a b s t r a c t Background: Homocysteine is a potential predictor for contrast-induced nephropathy (CIN). We aimed to compare homocysteine with cystatin C as pre-procedure predictors for CIN in patients undergoing coronary computed tomography angiography (CCTA). Methods: A total of 580 consecutive patients were enrolled. Concentrations of plasma homocysteine and serum cystatin C were measured before CCTA. CIN is defined as an elevation of creatinine by ≥ 25% or ≥ 0.5 mg/dl from baseline within 48 h. Receiver operating characteristic curves, Pearson correlation coefficients and logistic regression analysis were used to evaluate the efficiency of potential predictors. Results: Fifty-seven (9.83%) patients developed CIN. Concentrations of homocysteine (19.35 ± 4.32 μmol/l vs. 13.42 ± 3.96 μmol/l, p b 0.001) and cystatin C (1.20 ± 0.21 mg/dl vs. 0.99 ± 0.15 mg/dl, p b 0.001) increased significantly in CIN subjects. CIN was predicted by homocysteine (AUC 0.829, p b 0.001) and cystatin C (AUC 0.774, p b 0.001), while creatinine was not predictive. Both homocysteine and cystatin C had positive correlation with ΔCreatinine48h-0 (p b 0.001) and negative correlation with ΔeGFR48h-0 (p b 0.001). Regression analysis confirmed that increased baseline homocysteine [OR: 1.262 (1.123, 2.554), p b 0.001] and cystatin C [OR: 1.565 (1.380, 1.775), p b 0.001] were independent predictors for CIN. Conclusions: Homocysteine, with similar predictive value compared to cystatin C, was an independent biomarker for predicting CIN before CCTA examination. © 2015 Elsevier B.V. All rights reserved.
1. Introduction With the rapid advances in mechanical devices and interventional cardiology, a growing number of patients with suspected or known coronary artery disease are undergoing coronary computed tomography angiography (CCTA), conventional angiography and percutaneous coronary intervention in clinical practice. The extensive administration of contrast agents during these procedures has become a highly focused issue globally [1–3]. It is reported that contrast-induced nephropathy (CIN), a serious complication associated with contrast media administration, is the third most common cause of hospital-acquired acute kidney injury and accounts for about 11–12% of cases [4]. CIN is considered to be connected with increased morbidity, prolonged
⁎ Corresponding author at: Department of Cardiology, the Third Affiliated Hospital, Sun Yat-sen University, Tian-he Road, Guangzhou 510630, China. Tel.: +86 20 85252168; fax: +86 20 85252874. E-mail address: [email protected] (J. Liu).
http://dx.doi.org/10.1016/j.cca.2015.02.019 0009-8981/© 2015 Elsevier B.V. All rights reserved.
hospitalization, increased risk of complications, potential need for dialysis and increased mortality rate [4]. Conventionally, CIN is defined as an elevation of serum creatinine (sCr) of N 25% or ≥ 0.5 mg/dl (44 μmol/l) from baseline within 48 h after a diagnostic or interventional procedure in the absence of an alternative etiology [5–7]. However, sCr is not an ideal indicator for CIN, since it is secreted by renal tubules and influenced by muscle mass [8]. Therefore, more sensitive markers of CIN are desired. Recently, cystatin C, a cysteine protease inhibitor which has a constant production rate irrespective of muscle mass and a plasma concentration determined by glomerular filtration alone, has been well demonstrated to be superior to sCr for predicting CIN in a number of studies [4,9–13]. On the other hand, hyperhomocysteinemia is considered to be independently associated with a greater risk of CIN after angiography or angioplasty in two previous studies [14,15]. An explanation is that hyperhomocysteinemia induces oxidative stress, cellular apoptosis and endothelial dysfunction, which shares the proposed pathophysiologic mechanisms of CIN. However, the predicting value of plasma homocysteine comparing to serum cystatin C (an ideal predictor of CIN) has never been evaluated.
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
2. Materials and methods 2.1. Study population The current study was approved by the Institutional Review Board of the hospital. The written informed consents were obtained. From January, 2012 to July, 2014, consecutive inpatients receiving 320-slice CCTA (Aquilion ONE; Toshiba Medical Systems) for any reason were enrolled. The indications for performing CCTA were 1) evaluation of suspected coronary artery disease (depending on angina-like symptoms, cardiogenic syncope, elevated cardiovascular risk, abnormal echocardiogram, positive stress ECG test, positive myocardial perfusion scintigraphies), 2) determination of the patency of bypass grafts or stents, 3) evaluation of cardiomyopathy, and 4) preoperative evaluation for non-coronary surgery. Patients were excluded if they had any of the following conditions: 1) impaired renal function (eGFR b60 ml/min per 1.73 m2), 2) exposure to nephrotoxic drugs prior to or during the study period, 3) allergy to iodine-containing contrast medium, 4) contraindication for β-blockers, and 5) pregnancy. 2.2. Serum measurements Blood samples for evaluations of plasma homocysteine and serum cystatin C were collected at baseline, while samples for estimations of sCr were collected at 0, 24 and 48 h of CCTA. Blood samples were immediately sent to the central laboratory of the hospital, and analyzed on the Siemens ADVIA Centaur for measurement of homocysteine, or on the Hitachi 7180 clinical chemistry analyzer for measurement of cystatin C and sCr. Plasma homocysteine was measured using a competitive immunoassay with direct chemiluminescence detection. Serum cystatin C was measured via an immunoturbidimetric technique. sCr was estimated by Jaffe's method. Estimated glomerular filtration rate (eGFR) was calculated by the Modification of Diet in Renal Disease (MDRD) study equation [16]. Baseline concentrations of blood urea nitrogen (BUN), total cholesterol, triglyceride, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), uric acid, albumin and fasting glucose were also measured using ITACHI 7180 chemistry analyzer. Baseline hemoglobin was measured by the cyanmethemoglobin method. According to the definition of CIN relied on the changes of sCr within 48 h after CCTA, patients were assigned to either CIN(+) or CIN(−) group. 2.3. Coronary computed tomography angiography Before CCTA examination, 500 ml oral fluid was given to each patient. After CCTA, patients took unrestrictively oral fluid as much as possible. Oral metoprolol tablets (50–100 mg, Astrazeneca) were given to patients with a resting heart rate N 70 bpm. Isovue-370 (50–100 ml, Bracco Diagnostics) was injected intravenously at a flow rate of 6.0 ml s−1, followed by 20 ml saline flush at a flow rate of 4.0 ml s− 1. 320-slice CCTA was performed with 0.5-mm detector element, 350 ms of gantry rotation time, and up to 16 cm of coverage in Z direction. Tube voltages were set at 100–135 kV and the maximal tube currents were set at 400–580 mA. 2.4. Statistical analysis Statistical analysis was performed by using SPSS 18.0. Categorical variables were expressed as percentages, and continuous variables were presented as means ± SD. Differences of continuous variables between groups were evaluated by Student's t test, while differences of categorical variables were evaluated by Pearson chi-square test. A p b 0.05 was considered to be statistically significant, unless otherwise specified. Receiver operating characteristic (ROC) curves were drawn to detect the optimum cut-offs of parameters and their sensitivity and specificity in predicting CIN. The optimum cut-off is defined as the
87
point where one gets a maximum value of Youden's index (J = sensitivity + specificity − 1) in the ROC curve. Pearson correlations were used to explore the associations between potential biomarkers and the changes of sCr and eGFR after CCTA. In addition, a univariate logistic regression analysis was performed through potential predictive factors. And then, a multivariate logistic regression analysis was used to assess the correlation among the parameters which got a level of p b 0.05 in the univariate analysis.
3. Results 3.1. Patient characteristics A total of 580 eligible patients were enrolled. Fifty-seven (9.83%) of them developed CIN after the infusion of contrast medium, while 523 (90.17%) of them did not. Among patients developing CIN, 70.18% of them were males, which was higher than the percentage in the CIN(−) group. CIN(+) patients were significantly older than CIN(−) ones on average (67.2 ± 9.4 vs. 62.6 ± 10.9 y, p = 0.002). In addition, CIN(+) patients had larger portion of diabetes (71.9% vs. 57.9%, p = 0.041), higher total cholesterol level (139.0 ± 18.1 mg/dl vs. 132.4 ± 19.6 mg/dl, p = 0.016), and poor HbA1c level (6.84 ± 1.10% vs. 6.53 ± 1.04%, p = 0.031) than those of CIN(−) patients. However, the other baseline profiles showed no significant difference between CIN(+) and CIN(−) groups (Table 1).
3.2. Pre-CCTA concentrations of biomarkers At baseline (Fig. 1), the levels of sCr (0.94 ± 0.06 vs. 0.93 ± 0.09 mg/dl, p = 0.556), eGFR (80.86 ± 6.86 1.73 m2 vs. 83.35 ± 10.25 ml/min · 1.73 m2 , p = 0.085) and BUN (15.58 ± 2.61 vs. 14.97 ± 2.40 mg/dl, p = 0.074) were similar between CIN(+) and CIN(−) groups, whereas pre-CCTA concentrations of homocysteine (19.35 ± 4.32 vs. 13.42 ± 3.96 μmol/l, p b 0.001) and cystatin C (1.20 ± 0.21 vs. 0.99 ± 0.15 mg/dl, p b 0.001) in CIN(+) patients were significantly higher than those in CIN(−) subjects.
Table 1 Characteristics of eligible patients. Characteristics
CIN(+) (n = 57)
CIN(−) (n = 523)
p
Male/Female, n1/n2 Age, y Body mass index, kg/m2 Hypertension, n Diabetes, n Smoker, n Alcohol, n History of myocardial infarction, n Single-/double-/triple-vessel disease, n1/n2/n3 Contrast, ml Fluid intake after CCTA, ml Statin, n Hemoglobin, g/l Total cholesterol, mg/dl Triglyceride, mg/dl HDL-C, mg/dl LDL-C, mg/dl Albumin, g/l Uric acid, μmol/l Fasting glucose, mmol/l HbA1c, % Left ventricular ejection fraction, %
40/17 67.2 ± 9.4 23.01 ± 3.01 18 41 39 43 4
288/235 62.6 ± 10.9 23.44 ± 2.66 152 303 355 333 25
0.029a 0.002b NSb NSa 0.041a NSa NSa NSa
12/9/6
89/80/53
0NSa
74.3 ± 9.7 1240 ± 184 46 118.7 ± 12.8 139.0 ± 18.1 100.0 ± 14.6 31.5 ± 3.5 78.9 ± 8.4 36.4 ± 3.1 332.6 ± 64.7 7.1 ± 1.4 6.84 ± 1.10 59.5 ± 8.1
73.5 ± 8.1 1229 ± 200 458 120.2 ± 15.4 132.4 ± 19.6 102.2 ± 13.6 30.8 ± 3.49 77.5 ± 9.8 36.3 ± 3.4 347.1 ± 73.8 6.7 ± 1.5 6.53 ± 1.04 59.0 ± 7.7
NSb NSb NSa NSb 0.016b NSb NSb NSb NSb NSb NSb 0.031b NSb
a b
p value calculated by Pearson χ2 test. p value calculated by Student's t test.
88
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
Fig. 1. Baseline concentrations of homocysteine (a), cystatin C (b), sCr (c), eGFR (d) and BUN (e). Homocysteine and cystatin C were significantly higher in the CIN(+) group than the CIN(−) group, whereas sCr, eGFR and BUN had no differences.
3.3. ROC curves of potential biomarkers ROC curves of potential biomarkers for predicting CIN were drawn in Fig. 2. The areas under the curve (AUC) of homocysteine (AUC 0.829, 95% CI 0.771–0.886) and cystatin C (AUC 0.774, 95% CI 0.700–0.848) were significantly higher than those of sCr (AUC 0.569, 95% CI 0.502– 0.636), eGFR (AUC 0.571, 95% CI 0.486–0.656) and BUN (AUC 0.571, 95% CI 0.487–0.655). Thus, both pre-CCTA homocysteine (p b 0.001)
and cystatin C (p b 0.001) had predictive values for CIN, while sCr (p = 0.087), eGFR (p = 0.078) and BUN (p = 0.079) had no predictive values. At a optimum cut-off of 16.65 μmol/l, plasma homocysteine had a maximum sensitivity of 71.9% and specificity of 75.0% for predicting CIN. At an optimum cut-off of 1.05 mg/dl, serum cystatin C had a maximum sensitivity of 68.4% and specificity of 71.7% for predicting CIN. 3.4. Associations between potential biomarkers and the changes of sCr and eGFR As shown in Table 2, pre-CCTA plasma homocysteine had positive correlation with ΔsCr48h-0 (r = 0.356, p b 0.001) and negative correlation with ΔeGFR48h-0 (r = −0.307, p b 0.001). However, homocysteine had no correlations with ΔsCr24h-0 and ΔeGFR24h-0. Unlike that, preCCTA serum cystatin C had positive correlations with both ΔsCr24h-0 (r = 0.164, p b 0.001) and ΔsCr48h-0 (r = 0.405, p b 0.001), and negative correlations with both ΔeGFR24h-0 (r = − 0.163, p b 0.001) and ΔeGFR48h-0 (r = − 0.410, p b 0.001). Although baseline sCr, eGFR and BUN also had some correlations with the changes of sCr and eGFR at either 24 h or 48 h after CCTA, the correlation coefficients were relatively low (r = −0.197 to 0.185). 3.5. Logistic regression analysis
Fig. 2. The receiver operating characteristic (ROC) curves of baseline homocysteine, cystatin C, sCr, eGFR and BUN. Both homocysteine (AUC 0.829, p b 0.001; optimum cutoff 16.65 μmol/l, sensitivity 71.9%, specificity 75.0%) and cystatin C (AUC 0.774, p b 0.001; optimum cut-off 1.05 mg/dl, sensitivity 68.4%, specificity 71.7%), had predictive values for CIN, while sCr (AUC 0.569, p = 0.087), eGFR (AUC 0.571, p = 0.078) and BUN (AUC 0.571, p = 0.079) had none.
In the univariate logistic regression analysis, gender, age, total cholesterol, sCr, eGFR, homocysteine and cystatin C were found to be potentially correlated with the development of CIN (p b 0.05) (Table 3). As shown in Table 4, the multivariate logistic regression analysis confirmed that increased baseline homocysteine (odds ratio 1.262, 95% CI 1.123– 2.554, p b 0.001) and cystatin C (odds ratio 1.565, 95% CI 1.380–1.775, p b 0.001) were independent predictors for the development of CIN, while baseline sCr (odds ratio 0.987, 95% CI 0.280–1.989, p = 0.195) and eGFR (odds ratio 0.966, 95% CI 0.503–2.162, p = 0.251) were not independent predictors. In addition, males (odds ratio 1.363, 95% CI
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
89
Table 2 Pearson correlations. ΔsCr48h-0
ΔsCr24h-0
ΔeGFR24h-0
ΔeGFR48h-0
Biomarkers
Correlation
p
Correlation
p
Correlation
p
Correlation
p
Homocysteine Cystatin C sCr eGFR BUN
0.101 0.164 0.056 0.064 0.057
0.015a b0.001a NS NS NS
0.356 0.405 0.185 −0.041 0.106
b0.001a b0.001a b0.001a NS 0.011a
−0.079 −0.163 0.124 −0.103 −0.065
NS b0.001a 0.003a 0.021a NS
−0.307 −0.410 0.025 −0.197 −0.153
b0.001a b0.001a NS b0.001a b0.001a
a
p ≤ 0.05, indicating significant correlations between biomarkers and the changes of sCr or eGFR.
1.022–1.364, p = 0.005), elders (odds ratio 1.197, 95% CI 1.047–1.150, p b 0.001), and patients with higher total cholesterol levels (odds ratio 1.026, 95% CI 1.004–1.049, p = 0.023) seemed to have higher risks for developing CIN. 4. Discussion The present prospective study was the first “head-to-head” trial to compare plasma homocysteine with serum cystatin C for the prediction of CIN in patients undergoing CCTA. The authors found that both plasma homocysteine and serum cystatin C were independent biomarkers for pre-CCTA prediction of CIN. At respective optimum cut-offs (16.65 μmol/l for homocysteine, and 1.05 mg/dl for cystatin C), plasma homocysteine and serum cystatin C had similar sensitivity (71.9% vs. 68.4%) and specificity (75.0% vs. 71.7%) for predicting CIN. The occurrence of CIN in the present study was 9.83%, which is similar to those of previous trials [5,9,10]. Those patients with positive results on CCTA (narrowing of coronary lumen ≥50%) or high clinical probability of coronary artery disease tended to undergo further angiography and percutaneous coronary intervention in clinical routine, leading to the overexposure of contrast media, deterioration of renal function and adverse outcomes. Although several nonpharmacological and pharmacological approaches have been evaluated for the management of CIN, there is no definitive treatment available for CIN, making
Table 3 Univariate logistic regression analysis of potential factors determining CIN. Factors
Male Age Body mass index Hypertension Diabetes Smoker Alcohol Coronary artery disease Contrast volume Statin Hemoglobin Total cholesterol Triglyceride HDL-C LDL-C Albumin Fasting glucose Uric acid HbA1c Left ventricular ejection fraction Fluid intake after CCTA BUN Serum creatinine eGFR Cystatin C Homocysteine
Odds ratio
1.575 1.096 0.991 1.764 2.485 1.685 1.586 1.127 1.012 0.989 1.012 1.029 0.997 1.031 1.042 0.965 1.298 0.998 1.457 0.993 0.999 0.900 1.178 1.124 1.885 1.591
95% CI
p
Lower
Upper
1.013 1.037 0.797 0.629 0.934 0.535 0.542 0.450 0.959 0.746 0.981 1.003 0.962 0.909 0.995 0.850 0.942 0.991 0.954 0.940 0.997 0.740 1.076 1.035 1.048 1.387
2.394 1.158 1.232 4.950 6.613 5.305 4.645 2.820 1.086 1.355 1.043 1.055 1.034 1.170 1.091 1.097 1.788 1.004 2.226 1.048 1.002 1.096 2.048 1.222 4.173 1.826
0.049a 0.001a NS NS NS NS NS NS NS NS NS 0.028a NS NS NS NS NS NS NS NS NS NS 0.014a 0.006a b0.001a b0.001a
a Potential factors on the univariate analysis were added to multivariate logistic regression analysis for assessing the correlation among the parameters.
prevention the corner stone of management [17]. With the main goals of helping clinicians weight the risk of the exposure versus its benefit, risk scoring and stratification is a matter of concern for CIN prevention before contrast media administration. Recently, several risk scoring and stratification systems for CIN have been developed, such as the Bartholomew's score system, the CR4EATME3AD3 score, and the Mehran's score system [18–21]. These models were usually established by the conventional risk factors for CIN, including elderly age, impaired renal function, urgent coronary procedure, use of IABP, large contrast volume, left ventricular dysfunction, history of myocardial infarction, diabetes mellitus, hypertension, hypotension, peripheral vascular disease, anemia, hypoalbuminemia, dehydration, and cirrhosis. The present study confirmed that males, elders, and patients with higher total cholesterol levels seemed to have higher risks for the development of CIN. However, models based on conventional risk factors only were not sensitive and reliable enough to predictive CIN. One typical example is that CIN might develop even in subjects with normal sCr at baseline, which is an intrinsic flaw of sCr so-called “creatinine blind range” [22]. There is usually a 24–48 h delay between contrast exposure and the change in sCr. This delay makes sCr a late and unreliable indicator for the acute change of renal function [8]. Therefore, it is important to find more sensitive and accurate markers to predict CIN before the procedure. Fortunately, several biomarkers have been under evaluation, such as neutrophil gelatinase-associated lipocalin, cystatin C, interleukin-18, liver-type fatty acid-binding protein, and kidney injury molecule-1 [9,17,23,24]. Among them, cystatin C receives most interest in the searches for ideal predictors of CIN. In fact, cystatin C has been demonstrated to be an independent predictor for CIN in the present and previous studies [4,9–13]. Investigators found that CIN was predicted by baseline cystatin C independently, whereas baseline sCr were not predictive. There are several theoretical reasons why cystatin C is superior to sCr for predicting CIN [5,8,12,13,25]. sCr is a byproduct of muscle breakdown. It might cause the overestimation of eGFR in individuals with lower sCr generation due to loss of muscle mass. In addition, sCr can be actively secreted by renal tubules, which further reduces the value of sCr to estimate eGFR. Unlike that, cystatin C has a constant production rate irrespective of muscle mass, and a plasma concentration determined by glomerular filtration alone, which is not subject to tubular secretion.
Table 4 Multivariate logistic regression analysis of potential factors determining CIN. Factors
Male Age Total cholesterol sCr eGFR Cystatin C Homocysteine a
Odds ratio
1.363 1.097 1.026 0.987 0.966 1.565 1.262
95% CI
p
Lower
Upper
1.022 1.047 1.004 0.280 0.503 1.380 1.123
4.364 1.150 1.049 1.989 2.162 1.775 2.554
0.005a b0.001a 0.023a NS NS b0.001a b0.001a
p ≤ 0.05, indicating independent predictive factors for the development of CIN.
90
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91
On the contrary, plasma homocysteine receives less attention among the studies of CIN. Up to now, only 2 studies have investigated the relationship between homocysteine and the risk of CIN [14,15]. In a cohort of 572 patients who underwent percutaneous coronary intervention, Kim et al. found that the incidence of CIN was significantly greater in patients in the third homocysteine tertile (from lowest to highest, 4.7%, 7.3%, and 24.2%, p b 0.001). Hyperhomocysteinemia was an independent risk factor for CIN (odds ratio 1.70, 95% CI 1.07–2.71, p = 0.025) [15]. Similarly, Barbieri et al. investigated 876 patients with impaired renal function (eGFR b60 ml/min) who underwent coronary angiography and/or angioplasty, and found a significant relationship between homocysteine and CIN (odds ratio 1.68, 95% CI 1.09–2.59, p = 0.019) [14]. Therefore, homocysteine is a new potential biomarker for CIN, which deserves more attention in the near future. However, the predicting value of plasma homocysteine comparing to serum cystatin C, the ideal predictor of CIN, has never been evaluated. As the first attempt, the present study confirmed that plasma homocysteine, with similar sensitivity and specificity at the optimum cut-off compared to cystatin C, was an independent biomarker for the prediction of CIN after correction of the baseline confounding factors. Unlike Kim's or Barbier's studies, the current trial focused on patients receiving CCTA, a widespread non-invasive approach for evaluating coronary artery trees. It is expected to help a large number of patients weight the risk of CCTA versus its benefit. The link between homocysteine and the risk of CIN can be partly explained by the intrinsic pathophysiologic mechanisms they shared. Although the definite mechanism of CIN is not well-established, several mechanisms have been proposed: 1) direct toxicity of contrast media which is related to free radicals and oxidative stress; 2) apoptosis of renal cells induced by contrast media; and 3) renal medullary hypoxia [17,26,27]. Coincidentally, homocysteine is a potent excitatory neurotransmitter that binds to the N-methyl-D-aspartate receptor, leading to oxidative stress, cellular apoptosis, and endothelial dysfunction. Therefore, it is reasonable to hypothesize that hyperhomocysteinemia and CIN might have a causal relationship. However, the present observational study is unable to give a definite answer. N-acetylcysteine, an antioxidant which lowers homocysteine, has been shown to protect patients from CIN after coronary angiographic procedures [28,29]. The protective role of N-acetylcysteine is attributed to the inhibition of oxygen free radical production upon contrast media exposure. Unfortunately, none of the previous studies were primarily designed to lower homocysteine, and the change of homocysteine after N-acetylcysteine treatment had not been well evaluated. Therefore, the relationship between homocysteine-lowering therapy and CIN remains uncertain. Further prospective interventional study is needed in the near future to clarify whether homocysteine-lowering therapy among patients with hyperhomocysteinemia reduces the incidence of CIN or not. 4.1. Clinical implications The present study provides available data for the prediction of CIN in patients receiving CCTA, and is a good supplement to the previous risk score models. Based on the levels of serum cystatin C and plasma homocysteine, physicians can obtain more information for the prevention of CIN prior to CCTA, by using less contrast agent, sufficient hydration, and potential preventive agents. 4.2. Limitations It is a single center study, and trials on more mixed populations are warranted. The present study did not include patients with eGFR b60 ml/min per 1.73 m2, since these patients were not recommended to receive CCTA. Thus, the predictive value of sCr might be underestimated in the present study, which might limit its application in clinical practice. Furthermore, the authors only examined the changes
of biomarkers for 48 h, and long-term predictive values of biomarkers should be investigated further. 4.3. Conclusions CIN is a common potential complication after contrast administration in CCTA. Plasma homocysteine, with similar value compared to serum cystatin C, was an independent biomarker for pre-CCTA prediction of CIN. Studies on homocysteine-lowering therapy are warranted for reducing the incidence of CIN.
Abbreviations CIN sCr CCTA LVEF
contrast-induced nephropathy serum creatinine coronary computed tomography angiography left ventricular ejection fraction
References [1] EI-Hajjar M, Bashir I, Khan M, Min J, Torosoff M, DeLago A. Incidence of contrastinduced nephropathy in patients with chronic renal insufficiency undergoing multidetector computed tomographic angiography treated with preventive measures. Am J Cardiol 2008;102:353–6. [2] Ko B, Garcia S, Mithani S, Tholakanahalli V, Adabag S. Risk of acute kidney injury in patients who undergo coronary angiography and cardiac surgery in close succession. Eur Heart J 2012;33:2065–70. [3] Leoncini M, Toso A, Maioli M, Tropeano F, Villani S, Bellandi F. Early high-dose rosuvastatin for contrast-induced nephropathy prevention in acute coronary syndrome: results from the PRATO-ACS Study (Protective Effect of Rosuvastatin and Antiplatelet Therapy On contrast-induced acute kidney injury and myocardial damage in patients with Acute Coronary Syndrome). J Am Coll Cardiol 2014;63: 71–9. [4] Maliborski A, Zukowski P, Nowicki G, Bogusławska R. Contrast-induced nephropathy—a review of current literature and guidelines. Med Sci Monit 2011; 17:RA199–204. [5] Ribichini F, Gambaro G, Graziani MS, et al. Comparison of serum creatinine and cystatin C for early diagnosis of contrast-induced nephropathy after coronary angiography and interventions. Clin Chem 2012;58:458–64. [6] Aspelin P, Aubry P, Fransson SG, Strasser R, Willenbrock R, Berg KJ. Nephrotoxic effects in high-risk patients undergoing angiography. N Engl J Med 2003;348: 491–9. [7] Mohammed NM, Mahfouz A, Achkar K, Rafie IM, Hajar R. Contrast-induced nephropathy. Heart Views 2013;14:106–16. [8] Dalton RN. Serum creatinine and glomerular filtration rate: perception and reality. Clin Chem 2010;56:687–9. [9] Padhy M, Kaushik S, Girish MP, Mohapatra S, Shah S, Koner BC. Serum neutrophil gelatinase associated lipocalin (NGAL) and cystatin C as early predictors of contrast-induced acute kidney injury in patients undergoing percutaneous coronary intervention. Clin Chim Acta 2014;435:48–52. [10] Alharazy SM, Kong N, Saidin R, et al. Serum neutrophil gelatinase-associated lipocalin and cystatin C are early biomarkers of contrast-induced nephropathy after coronary angiography in patients with chronic kidney disease. Angiology 2014;65:436–42. [11] Wacker-Gußmann A, Bühren K, Schultheiss C, et al. Prediction of contrast-induced nephropathy in patients with serum creatinine levels in the upper normal range by cystatin C: a prospective study in 374 patients. AJR Am J Roentgenol 2014;202: 452–8. [12] Nozue T, Michishita I, Mizuguchi I. Predictive value of serum cystatin C, β2microglobulin, and urinary liver-type fatty acid-binding protein on the development of contrast-induced nephropathy. Cardiovasc Interv Ther 2010;25:85–90. [13] Briguori C, Visconti G, Rivera NV, et al. Cystatin C and contrast-induced acute kidney injury. Circulation 2010;121:2117–22. [14] Barbieri L, Verdoia M, Schaffer A, et al. Elevated homocysteine and the risk of contrastinduced nephropathy: a cohort study. Angiology 2014 [pii: 0003319714533401, Epub ahead of print]. [15] Kim SJ, Choi D, Ko YG, et al. Relation of homocysteinemia to contrast-induced nephropathy in patients undergoing percutaneous coronary intervention. Am J Cardiol 2011;108:1086–91. [16] Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of diet in renal disease study group. Ann Intern Med 1999; 130:461–70. [17] Mohammed NM, Mahfouz A, Achkar K, Rafie IM, Hajar R. Contrast-induced Nephropathy. Heart Views 2013;14:106–16. [18] Bartholomew BA, Harjai KJ, Dukkipati S, et al. Impact of nephropathy after percutaneous coronary intervention and a method for risk stratification. Am J Cardiol 2004; 93:1515–9.
S. Li et al. / Clinica Chimica Acta 444 (2015) 86–91 [19] Fu N, Li X, Yang S, et al. Risk score for the prediction of contrast induced nephropathy in elderly patients undergoing percutaneous coronary intervention. Angiology 2013; 64:188–94. [20] Mehran R, Aymong ED, Nikolsky E, et al. A simple risk score for prediction of contrast-induced nephropathy after percutaneous coronary intervention: development and initial validation. J Am Coll Cardiol 2004;44:1393–9. [21] Tziakas D, Chalikias G, Stakos D, et al. Development of an easily applicable risk score model for contrast-induced nephropathy prediction after percutaneous coronary intervention: a novel approach tailored to current practice. Int J Cardiol 2013;163: 46–55. [22] Takeuchi T, Isobe S, Sato K, et al. Cystatin C: a possible sensitive marker for detecting potential kidney injury after computed tomography coronary angiography. J Comput Assist Tomogr 2011;35:240–5. [23] Ling W, Zhaohui N, Ben H, et al. Urinary IL-18 and NGAL as early predictive biomarkers in contrast-induced nephropathy after coronary angiography. Nephron Clin Pract 2008;108:c176–81.
91
[24] Nogare AL, Dalpiaz T, Veronese FJ, Gonçalves LF, Manfro RC. Noninvasive analyses of kidney injury molecule-1 messenger RNA in kidney transplant recipients with graft dysfunction. Transplant Proc 2012;44:2297–9. [25] Hatano M. Clinical study of tubular creatinine secretion in renal dysfunction. Jpn J Nephrol 1991;33:1097–104. [26] Romano G, Briguori C, Quintavalle C, et al. Contrast agents and renal cell apoptosis. Eur Heart J 2008;29:2569–76. [27] Hizoh I, Haller C. Radiocontrast-induced renal tubular cell apoptosis: hypertonic versus oxidative stress. Invest Radiol 2002;37:428–34. [28] Kay J, Chow WH, Chan TM, et al. Acetylcysteine for prevention of acute deterioration of renal function following elective coronary angiography and intervention: a randomized controlled trial. JAMA 2003;289:553–8. [29] Marenzi G, Assanelli E, Marana I, et al. N-acetylcysteine and contrast-induced nephropathy in primary angioplasty. N Engl J Med 2006;354:2773–82.
