International Journal of Cardiology 174 (2014) 37–42
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Female gender and contrast-induced nephropathy in primary percutaneous intervention for ST-segment elevation myocardial infarction Stefano Lucreziotti a,⁎,1, Marco Centola a,1, Diego Salerno-Uriarte a,1, Giorgio Ponticelli a,1, Pier Maria Battezzati b,1, Diego Castini a,1, Carlo Sponzilli a,1, Federico Lombardi c,1 a b c
Cardiologia, Ospedale San Paolo, Dipartimento di Scienze della Salute, University of Milan, Italy Divisione di Medicina VI, Ospedale San Paolo, Dipartimento di Scienze della Salute, University of Milan, Italy U.O.C. di Malattie Cardiovascolari, Fondazione IRCCS, Ospedale Maggiore Policlinico, Dipartimento di Scienze Cliniche e di Comunità, University of Milan, Italy
a r t i c l e
i n f o
Article history: Received 11 February 2013 Received in revised form 10 October 2013 Accepted 14 March 2014 Available online 20 March 2014 Keywords: Contrast-induced nephropathy Coronary angioplasty Female gender Mortality Acute myocardial infarction
a b s t r a c t Background: Patients undergoing primary percutaneous coronary intervention (PCI) are at high risk for contrastinduced nephropathy (CIN), a complication that has been demonstrated to negatively affect outcomes. It has been suggested that, when compared to males, female patients present higher incidence of CIN and higher mortality after primary PCI. However, the specific role of gender in this setting remains ill-defined given its complex interplay with several co-morbidities and clinical characteristics. We investigated the relationship of patients' variables, including gender, with CIN and mortality after primary PCI. Methods: In a single center study in 323 consecutive patients undergoing primary PCI, the development of CIN and mortality during an 18-month median follow-up period was assessed. CIN was defined as an increase in serum creatinine (≥25% or ≥0.5 mg/dl) from baseline occurring at any time during the first 3 post-procedural days. Results: CIN occurred in 23 female and 26 male patients (25.0% vs 11.2%, p = 0.003), while cumulative mortality was 10.6%. Women presented unfavorable basal characteristics and underwent myocardial reperfusion less quickly. At multivariable analysis, reduced left ventricular ejection fraction (LVEF) (odds ratio [OR] 7.32 95% confidence interval [CI]: 2.60–21, p b 0.001) and female gender (OR 2.49 95%CI 1.22–5.07, p = 0.01) predicted CIN, whereas the occurrence of CIN (hazard ratio [HR] 3.65 95%CI 1.55–8.59, p = 0.003) and a Mehran risk score (MRS) ≥6 (HR 1.76 95%CI 1.13–2.74, p = 0.01) independently predicted long-term mortality. Conclusions: After primary PCI, female gender and LVEF are associated with an increased risk of CIN, whereas MRS and development of CIN predict long-term mortality. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Primary percutaneous coronary intervention (PCI) is the best reperfusion strategy in patients presenting with ST-segment elevation myocardial infarction (STEMI) [1,2]. When compared to elective procedures, primary PCI is associated with a higher incidence of contrast-induced nephropathy (CIN), a complication which is associated with increased in-hospital and long-term morbidity and mortality [3].
⁎ Corresponding author at: Cardiologia, Ospedale San Paolo, Via A. di Rudinì 8, 20142 Milan, Italy. Tel./fax: +39 02 50323145. E-mail address: [email protected] (S. Lucreziotti). 1 This author takes responsibility of all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2014.03.087 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
It would be therefore useful and desirable to promptly detect patients at high risk for CIN because they could be carefully monitored and treated with adequate prophylactic strategy. Patients prone to develop CIN are those with older age, hemodynamic instability, long symptom to reperfusion time, and anterior infarction [3]. Although it has been suggested that female gender exposes patients to higher risk for CIN, no conclusive evidence is available, particularly in the setting of primary PCI [3–7]. Uncertainty also exists regarding gender differences in survival after primary PCI: women have a worse short- and long-term prognosis than men, even though this difference seems to be largely influenced by comorbidities and clinical variables, which characterize female gender [8–10]. The aim of the present study was to investigate the correlation between female gender and CIN in a cohort of STEMI patients undergoing primary PCI. The role of female gender and CIN in predicting mortality was also evaluated.
38
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
2. Methods 2.1. Patients Data on STEMI patients admitted from January 2009 to December 2011 were prospectively entered and retrospectively analysed. We enrolled all consecutive patients who presented to the emergency department of our hospital within 12 h (24 h in the presence of cardiogenic shock or clinical/electrocardiographic evidence of ongoing ischaemia) from the onset of symptoms with a ST-segment elevation electrocardiographic pattern in two or more contiguous leads, or left bundle branch block. Exclusion criteria were: cardiac arrest requiring a prolonged resuscitation (N30 min), ongoing dialytic treatment, and malignancy with a life expectancy of less than 6 months. All patients were followed up by means of a hospital visit at 6 months and every subsequent year or by telephone interviews. This investigation was designed as an observational study and it conforms to the principles outlined in the Declaration of Helsinki; the Research Ethics Board of the San Paolo Hospital and the University of Milan, Italy approved the study protocol. An informed consent was obtained from all patients. 2.2. Coronary angioplasty and medical therapy Before undergoing coronary angiography, all patients underwent a rapid clinical assessment with blood sampling and 12-lead ECG recording. Medical therapy and modality of revascularization were left at the discretion of the coronary care and interventional cardiologists. Prophylaxis for CIN was not routinely administered. All patients received iso-osmolar, nonionic contrast agent iodixanol. Primary PCI was performed by an interventional team on a 24-hour basis with a standard technique via femoral or radial approach. Periprocedural antithrombotic therapy was prescribed according to the current guidelines [1]. Briefly, a loading dose of aspirin (300 to 500 mg) and clopidogrel (300 or 600 mg) or prasugrel (60 mg) was administered before the procedure, while anticoagulant therapy was started after arterial sheath insertion. Heparin was administered at the doses of 100 U/kg or 70 U/kg, if an additional therapy with abciximab was planned. Additional heparin boluses were eventually added in order to maintain the activated clotting time N300 s, or between 200 and 250 if abciximab was used. Intravenous bivalirudin (0.75 mg/kg bolus and 1.75 mg/kg/h infusion for the duration of the procedure) was an alternative anticoagulant therapy. An early removal of arterial sheath was encouraged. In patients with multivessel disease, only the infarctrelated artery was treated during the acute phase. Delayed percutaneous or surgical revascularization was eventually performed in a later time (after at least 4 days, if the patients were clinically stable) in case of critical stenosis of one or more coronary arteries supplying a significant amount of jeopardized myocardium. 2.3. Other diagnostic techniques and definitions Echocardiographic assessment of the left ventricular ejection fraction (LVEF) was performed within 24 h after primary PCI. Values b40% were considered to reflect a reduced LVEF. Serum creatinine was measured at hospital admission and every day for the following 3 days. Serum creatinine levels ≥1.2 mg/dl were considered elevated. Estimated glomerular filtration rate (eGFR) was calculated according to the CKD-EPI equation [11]. CIN was defined as an increase ≥25% or ≥0.5 mg/dl (44 mmol/l) in serum creatinine from the baseline values within the first 72 h after the index procedure [12]. Major bleeding was defined according to the ‘Thrombolysis in Myocardial Infarction’ (TIMI) criteria [13]. A hemoglobin value b13 g/dl for men and b12 g/dl for women defined anemia [14]. Time to reperfusion was calculated as an interval between the symptom onset and the first balloon inflation. The risk for CIN was calculated in relation to the Mehran risk score (MRS) [15]. According to value distribution, patients were classified as low or medium-to-high risk when MRS was b or ≥6.
Death from any cause was the other major clinical outcome of the study. To summarize time to death data, death rates from any cause were expressed as the number of patients dying per 100 person-months, with death rates ratios and 95% CI reported to contrast risk groups. Cumulative proportions of patients dying were estimated by means of the Kaplan and Meier approach using the date of presentation to the Emergency Department as the starting point. Univariable analyses of risk factors for death were performed using the log-rank test. The proportional hazard Cox model was used to assess within the framework of multivariable analysis the effects of risk factors associated (p b 0.10) with death at univariable analysis while simultaneously adjusting for the same set of predefined variables used in the models predicting development of CIN. Likelihood ratio tests were used to test the joint effect of adding, or omitting, specific groups of variables on the predictive value of a Cox model. To avoid problems inherent in the categorization of continuous variables, noncategorical variables were preferentially introduced in multivariable models as continuous variables, unless categorization was strongly suggested by widespread prior use. For example, the MRS was introduced as categorical variable with score values grouped in 4 classes (MRS b6, 6–10, 11–16, or N16). In order to facilitate interpretation and visualization of results, in some instances analyses were also done arranging patients in two groups, i.e., those with low (b6) or moderate-to-high (≥6) MRS values. Statistical comparisons were performed using Stata Statistical Software (version 11, Stata Corporation, College Station, TX). All analyses were two-sided.
3. Results A total of 332 consecutive patients were considered. Nine patients were excluded from the study for the following reasons: dialysis (2 patients), coronary anatomy not suitable for PCI (4 patients), and death within the first 24 h after PCI (3 patients). Hence, 323 patients were included in the analysis. Follow-up was completed in 322 (99.7%) patients. Median follow-up period was 18 months (interquartile range 9.4–29.7 months) with an overall duration of 522 patient-years. 3.1. Patient characteristics and incidence of CIN Major clinical and procedural characteristics of the whole patients' group and of those who did or did not develop CIN are reported in Table 1. Seventy-nine (24.5%) patients were older than 75 years, and 55 (17%) had diabetes. Only a relatively small amount (9.3%) of subjects presented a reduced LVEF. Most patients were in Killip class 1 (87%) and in the low MRS class (64.3%). The left anterior coronary artery was the infarct-related vessel in 150 patients (46.4%). A multivessel disease was detected in 193 (59.8%) patients. The rate of major in-hospital bleeding was 4.7%. The mean value of creatinine was 1.0 ± 0.5 mg/dl. Serum creatinine levels were elevated in 56 (17.3%) patients. CIN was detected in 49 (15.2%) subjects, with a significantly (p = 0.003) higher incidence in women (25.0%) than in men (11.3%). 3.2. Predictors of CIN
2.4. Statistical analysis Continuous variables are summarized as mean ± SD values. The significance of differences between groups was assessed by the Mann–Whitney test, or by the Kruskal–Wallis test when more than two groups were concerned. Differences in proportions were tested using Fisher's exact test or chi-square statistics. The major outcomes of this study were development of CIN and death from any cause. Risk ratios (RR) and their 95% confidence intervals (CI) along with p values were used to describe the risk of developing CIN relative to a reference category. Mantel–Haenszel tests for linear trends were used to assess the presence of trends across different patients' categories or different levels of exposure. Logistic regression analysis was used to adjust risks, expressed as odd ratios (OR), while simultaneously adjusting for the effect of other variables under study. As stipulated in the study protocol, variables associated (p b 0.10) with the two major study outcome variables associated (p b 0.10) with development of CIN at univariable analysis were tested in multivariable models adjusted for the following variables: systolic blood pressure, hypertension, family history of cardiovascular events, anterior myocardial infarction, time to reperfusion, and multivessel coronary disease. Such variables were introduced regardless of their significance level in view of their potential role as confounding factors. Because MRS is a validated risk score for CIN and clinical outcome after PCI [7, 15], we did not include in multivariable models those variables concurring to the calculation of MRS and those directly associated with them, i.e., we did not include age, serum hemoglobin, diabetes, contrast media volume, serum creatinine and eGFR to avoid overfitting model or multicollinearity problems.
As shown in Table 1, CIN was more frequent in women and in patients with advanced age, reduced LVEF, impaired renal function and higher MRS. Mean MRS was 6.5 ± 4.8 in the patients who developed CIN compared with 5.2 ± 4.6 (p = 0.07) in those who did not develop the complication. The difference in BMI was of borderline significance, and largely depended on the different distribution of CIN between genders. When the distribution of BMI values into gender-specific quintiles was compared between patients according to development of CIN, such difference fell well below the significance level (p = 0.81). A similar result was found when BMI values were compared after adjustment for gender by logistic regression analysis (p = 0.37). The amount of contrast medium, the frequency of major bleeding, the basal glucose level, the electrocardiographic location of myocardial infarction and the Killip class were not significantly different between patients who developed or did not develop CIN. There were some relevant gender differences in basal and procedural risk profile (Table 2): women were older, and presented a higher MRS and a lower renal function. Mean MRS was 7.3 ± 5.0 in female patients compared with 4.7 ± 4.3
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
39
Table 1 Clinical and procedural characteristics of patients developing CIN. Variable
All patients (n = 323)
Patients with CIN (n = 49)
Patients without CIN (n = 274)
p value
Age, years Age ≥75 years Women Active smoking Diabetes mellitus Family history of cardiovascular disease Hypercholesterolemia Systolic blood pressure, mm Hg Hypertension BMI, kg/m2 LVEF b40% Anterior myocardial infarction Killip class 2–4 Multivessel coronary disease In-hospital major bleeding Time to reperfusion, min Contrast volume, ml Contrast volume N300 ml Hemoglobin, g/dl Glucose, mg/dl Creatinine, mg/dl eGFR, ml/min/1.73 m2 eGFR b60 ml/min/1.73 m2 MRS class b6 6–10 11–16 N16 MRS ≥6
65.3 ± 13.0 79 (24.5) 92 (28.5) 138 (42.7) 55 (17.0) 55 (17.0) 107 (33.1) 144 ± 27 169 (52.3) 26.1 ± 3.6 30 (9.3) 146 (45.2) 42 (13.0) 193 (59.8) 15 (4.7) 302 ± 320 259 ± 101 70 (21.7) 14.3 ± 1.8 159 ± 69 1.0 ± 0.5 76.7 ± 23.0 77 (23.8)
69.3 ± 14.7 18 (36.7) 23 (46.9) 20 (40.8) 10 (20.4) 6 (12.2) 12 (24.5) 148 ± 24 24 (49.0) 25.4 ± 3.8 13 (26.5) 24 (49.0) 9 (18.4) 27 (55.1) 5 (10.2) 341 ± 324 245 ± 85 7 (14.3) 14.0 ± 2.0 175 ± 85 1.2 ± 1.0 72.1 ± 31.6 20 (40.8)
64.6 ± 12.6 61 (22.3) 69 (25.2) 118 (43.1) 45 (16.4) 49 (17.9) 95 (34.7) 143 ± 27 145 (52.9) 26.2 ± 3.6 17 (6.2) 122 (44.5) 33 (12.0) 166 (60.6) 10 (3.7) 295 ± 320 262 ± 104 63 (23.0) 14.4 ± 1.8 156 ± 65 1.0 ± 0.4 77.5 ± 21.1 57 (20.8)
0.03 0.05 0.003 0.88 0.54 0.41 0.19 0.53 0.64 0.09 b0.001 0.64 0.25 0.53 0.06 0.12 0.4 0.19 0.11 0.17 0.28 0.58 0.006
207 (64.3) 69 (21.4) 36 (11.2) 10 (3.1) 115 (35.7)
23 (47.9) 14 (29.2) 9 (18.8) 2 (4.2) 25 (52.1)
184 (67.2) 55 (20.1) 27 (9.9) 8 (2.9) 90 (32.9)
0.07
0.01
Values are n (%) or mean ± SD, unless otherwise specified. CIN; contrast-induced nephropathy, BMI; body mass index, LVEF; left ventricular ejection fraction, eGFR; estimated glomerular filtration rate, MRS; Mehran risk score.
(p b 0.001) in males. In addition, women were treated with a considerable time-delay compared with males. On the other hand, males were more frequently, smokers. We have found a significant (p b 0.001) trend toward increasing risk of CIN across different gender and LVEF categories, with a strong interaction between these two variables (Fig. 1). At multivariable analysis (Table 3), female gender (OR 2.49 95%CI 1.22–5.07, p = 0.01) and reduced LVEF (OR 7.32 95%CI 2.60– 21, p b 0.001) were accompanied with significantly increased risk of CIN. In contrast, we did not observe any statistical association between MRS, in-hospital bleeding and CIN.
3.3. In-hospital and long-term mortality During hospital stay, 17 (5.3%) deaths occurred in the study population. In-hospital mortality was higher in patients developing CIN than in patients without this complication (20.4% vs 2.6%, RR 7.99 95%CI 3.19–20, p b 0.001), in women (9.8% vs 3.5%, RR 2.82 95%CI 1.12–7.10, p = 0.02), and in subjects with reduced LVEF (36.7% vs 2.0%, RR 17 95%CI 7.12–45, p b 0.001). When MRS was categorized into two classes, mortality was 11.3% in patients showing moderate-to-high values and 1.9% in those showing an MRS b 6 (RR 5.85 95%CI 1.95–18, p b 0.001).
Table 2 Clinical and procedural characteristics according to sex. Variable
Women (n = 92)
Men (n = 231)
p value
Age, years Active smoking Diabetes mellitus Family history of cardiovascular disease Hypercholesterolemia Hypertension LVEF b40% Killip class 2–4 Anterior myocardial infarction Time to reperfusion, min Contrast volume, ml MRS ≥6 Hemoglobin, g/dl Creatinine, mg/dl eGFR, ml/min/1.73 m2 eGFR b60 ml/min/1.73 m2 Glucose, mg/dl
72.0 ± 13.5 25 (27.2) 19 (20.7) 15 (16.3) 33 (35.9) 46 (50.0) 13 (14.1) 16 (17.4) 40 (43.5) 363 ± 357 253 ± 86 51 (56.0) 13.1 ± 1.5 1.0 ± 0.8 70.0 ± 26.1 36 (39.1) 161.1 ± 69.4
62.7 ± 11.8 113 (48.9) 36 (15.6) 40 (17.3) 74 (32.0) 123 (53.3) 17 (7.4) 26 (11.3) 106 (45.9) 278 ± 302 262 ± 107 64 (27.7) 14.8 ± 1.7 1.1 ± 0.4 79.4 ± 21.1 41 (17.7) 159.0 ± 69.0
b0.001 b0.001 0.33 0.87 0.52 0.62 0.09 0.15 0.71 0.002 0.81 b0.001 b0.001 b0.001 0.003 b0.001 0.68
Values are n (%) or mean ± SD, unless otherwise specified. BMI; body mass index, LVEF; left ventricular ejection fraction, MRS; Mehran risk score, eGFR; estimated glomerular filtration rate.
40
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42 Table 4 Factors associated with long-term-mortality at univariable analysis.
Fig. 1. Relative risk of CIN according to gender and LVEF. Male patients with LVEF ≥40% are the reference group, i.e., risk ratio is 1. Vertical bars denote 95% CI for the risk ratios; p values refer to the comparison with reference group. Mantel–Haenszel test for trend across the four categories: p b 0.001. CIN; contrast-induced nephropathy, LVEF; left ventricular ejection fraction, CI; confidence interval.
During the 18 month follow-up, 33 (10.6%) patients died. Clinical variables potentially associated with mortality were evaluated with univariable and multivariable analysis (Tables 4, 5). The multivariable analysis showed that only a moderate-to-high MRS and the occurrence of CIN were significantly associated with the risk of death, whereas the relationships between in-hospital major bleeding, reduced LVEF, female gender and long-term mortality were not statistically significant. Interestingly, there was a significant trend (p b 0.001) toward increasing mortality across the groups defined by moderate-to-high MRS and development of CIN (Fig. 2). 4. Discussion The main finding of the present study was that female gender was an independent predictor of CIN in STEMI patients treated with primary PCI. CIN development and high MRS were associated with increased mortality independently of gender. Previous studies have shown that the risk for CIN was higher in women undergoing elective or urgent PCI [5,6,16,17], but a correlation between female gender and CIN has never been reported in a selected population of STEMI patients treated by primary PCI [3,7]. A reduced glomerular volume in women compared to their male counterpart has been proposed as a possible mechanism to explain a greater susceptibility to acute renal failure [18]. It cannot be ruled out, however, that the association between female gender and CIN rather to be causal might be secondary to the higher prevalence of comorbidities that characterize female gender or, finally, to older age in our female population.
Table 3 Predictive factors for CIN at multivariable analysis. Variable
Odds ratio
95% CI
p value
LVEF b40% Female gender MRS class In-hospital major bleeding
7.32 2.49 1.02 2.61
2.60–21 1.22–5.07 0.63–1.64 0.70–9.77
b0.001 0.01 0.95 0.15
CIN; contrast induced nephropathy, OR; odds ratio, CI; confidence interval, LVEF; left ventricular ejection fraction, MRS; Mehran risk score.
Variable
Death rate
Death rate ratio
95% CI
p value
Age ≥75 years CIN Female gender Diabetes mellitus Hypertension LVEF b40% Anterior myocardial infarction Killip class 2–4 Multivessel coronary disease In-hospital major bleeding Time to reperfusiona Contrast volume N300 ml Anemia High Glucose levelb eGFR b60 ml/min/1.73 m2 MRS class b6 6–10 1–16 ≥16
1.93 2.05 1.06 0.75 0.60 3.79 0.62 1.94 0.62 2.70 0.48 0.55 1.13 0.67 2.24
2.82 6.03 3.07 1.55 1.37 11.23 1.36 5.41 1.62 5.86 0.88 1.05 3.00 1.37 12.93
4.49–23 2.80–13 1.46–6.50 0.61–3.56 0.65–3.00 5.13–23 0.65–2.89 2.47–11 0.73–3.97 1.77–15 0.30–2.19 0.41–2.41 1.39–6.31 0.56–3.06 5.80–32
b0.001 b0.001 0.001 0.26 0.31 b0.001 0.43 b0.001 0.13 0.001 0.79 0.75 0.002 0.27 b0.001
0.16 0.93 1.92 4.08
1 5.88 12.10 25.77
2.08–18 4.16–37 6.45–94
b0.001
CI; confidence interval, CIN; contrast-induced nephropathy, BMI; body mass index, LVEF; left ventricular ejection fraction, eGFR; estimated glomerular filtration rate, MRS; Mehran risk score. a Values in the higher quintile of the distribution of time to reperfusion. Cut-off level is 375 min. b Values in the higher quintile of the distribution of serum glucose level. Cut-off level is 190 mg/dl.
The new finding of an association between female gender and CIN is of relevant clinical and physiopathological significance but must be interpreted by considering some methodological features that characterize our study in comparison with previous ones. First, unlike previous STEMI studies [3,7], periprocedural hydration was not routinely administered in our study. Our population was therefore exposed to a higher risk of CIN due to the lack of prophylactic hydration that appears be more effective in women than in men [19]. Second, we determined CIN according to the recently validated CKD-EPI equation [11], at variance with previous studies where a different formula was used [3,7]. Regarding this methodological aspect, it has been reported that the CKD-EPI may provide a better evaluation of renal function than the MDRD formula, especially when measuring eGFR in a population that consists of patients with a preserved renal function [11]. Third, we included patients with cardiogenic shock who have been previously excluded in earlier studies [7]. Fourth, we defined CIN in relation to either absolute or percent increase in serum creatinine, whereas, for example, Marenzi et al. only considered absolute increases of this parameter [3]. Nevertheless, the incidence of CIN observed in our study was similar to that previously reported thus making unlikely that these methodological aspects might have played a major role. Finally, we also observed that a reduced LVEF was associated with an increased risk of CIN, whereas other traditional risk factors including the amount of injected contrast medium did not predict CIN development. This negative finding, which is in disagreement with previous reports [3,16], may Table 5 Predictors of long-term mortality at multivariable analysis. Variable
Hazard ratio
95% CI
p value
MRS class a CIN LVEF b40% In-hospital major bleeding Female gender
1.76 3.65 2.41 1.52 1.71
1.13–2.74 1.55–8.59 0.87–6.71 0.46–4.99 0.82–3.59
0.012 b0.003 0.09 0.49 0.15
CI; confidence interval, MRS; Mehran risk score, CIN; contrast-induced nephropathy, LVEF; left ventricular ejection fraction. a MRS was categorized in four classes. The reported hazard ratio and 95% CI are interpreted as the estimated hazard associated to a one-unit increase in MRS class.
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
41
Fig. 2. Kaplan–Meier estimates of mortality according to MRS and development of CIN. Patients are categorized based on the presence of a low (values b6) or a moderate-to-high (values ≥6) MRS and according to whether they did or did not develop CIN. MRS; Mehran risk score, CIN; contrast-induced nephropathy.
have been explained by the fact that in our center a selective strategy of revascularization limited to the infarct-related artery was adopted in all STEMI patients. Hence, the relatively low and less variable amount of injected contrast medium could account for the lack of association between this parameter and risk of CIN. Considerable debate exists whether gender has an independent prognostic role after STEMI. The present study shows that, after adjustment for other variables including CIN and MRS, long-term mortality after primary PCI was not higher in women than in men. These findings are consistent with the results of previous trials and registries [8–10, 20–22]. For example, a large Swedish registry has recently demonstrated a relevant gender gap regarding treatment appropriateness and mortality in patients with STEMI [20]; nevertheless, one year mortality was not higher in women when adherence to the treatment guidelines was considered in the multivariable analysis. Similar results were also observed in a large multicenter registry that also reported the presence of a significant difference in the rate of reperfusion treatment [21]. A lack of gender effect on long-term mortality was also found in two single center registries and in one multicenter, randomized, controlled trial, which, as in our study, enrolled only STEMI patients treated with primary PCI [8–10]. All these data suggest that women still receive suboptimal treatment and are more exposed to hard events after hospital discharge in the era of widespread use of reperfusion therapy and potent antithrombotic therapies. In our study, for example, women undergoing primary PCI presented a less favorable basal clinical profile and a longer reperfusion time. In addition, we observed that CIN development accounted for the higher in-hospital and overall mortality, a finding that further confirms the negative prognostic impact of this complication. The capability of MRS to predict mortality as well as major adverse cardiovascular and cerebrovascular events in STEMI patients treated with primary PCI has been previously reported [7]. We also found that MRS was clinically useful in predicting long-term mortality in this clinical context, thus confirming the predicting value for hard cardiovascular events of what has been considered a simple “renal” risk score. Quite surprisingly, we were unable to detect an association between MRS and CIN that has been previously proposed and validated in a large series of patients who underwent PCI, excluding those with an acute
myocardial infarction [15]. The different patients' subset, the inclusion of gender among the variables tested by multivariable analysis, the relative small sample size and the inhomogeneous distribution of patients in the four classes of MRS (Table 1) may account for this negative finding. Finally, it must be pointed out that we did not assess the independent prognostic value of those clinical or laboratory variables (for example, age and creatinine plasma levels) included in MRS.
5. Limitations This study was of smaller size compared with some of the available ones: this may have limited the statistical power to detect possible additional associations. It should be noted, however, that we enrolled an unselected population of all comers for STEMI who have been uniformly treated with primary PCI, as in the everyday clinical practice. This is a single-center study but with the advantage of a more homogeneous patients' management. We can't exclude that in some patients factors other than contrast medium including renal embolisms or dehydration might have contributed to the raise in plasma creatinine levels. Due to the limited number of deaths, only total mortality was considered. Finally, not all variables potentially affecting prognosis were evaluated in the present study.
6. Conclusions In patients undergoing primary PCI, female gender and reduced LVEF are associated with CIN risk. Both CIN and MRS, but not female gender, are independent predictors of long-term mortality at multivariable analysis.
References [1] Steg PG, James SK, Atar D, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: the Task Force on the management of ST-segment elevation acute myocardial infarction of the European Society of Cardiology (ESC). Eur Heart J 2012;33:2569–619.
42
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
[2] Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13–20. [3] Marenzi G, Lauri G, Assanelli E, et al. Contrast-induced nephropathy in patients undergoing primary angioplasty for acute myocardial infarction. J Am Coll Cardiol 2004;44:1780–5. [4] Mueller C, Buerkle G, Perruchoud AP, Buettner HJ. Female sex and risk of contrast nephropathy after percutaneous coronary intervention. Can J Cardiol 2004;20:505–9. [5] Sidhu RB, Brown JR, Robb JF, et al. Interaction of gender and age on post cardiac catheterization contrast-induced acute kidney injury. Am J Cardiol 2008;102:1482–6. [6] Thompson CA, Kaplan AV, Friedman BJ, et al. Gender-based differences of percutaneous coronary intervention in the drug-eluting stent era. Catheter Cardiovasc Interv 2006;67:25–31. [7] Sgura FA, Bertelli L, Monopoli D, et al. Mehran contrast-induced nephropathy risk score predicts short- and long-term clinical outcomes in patients with STelevation–myocardial infarction. Circ Cardiovasc Interv 2010;3:491–8. [8] Lansky AJ, Pietras C, Costa RA, et al. Gender differences in outcomes after primary angioplasty versus primary stenting with and without abciximab for acute myocardial infarction: results of the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial. Circulation 2005;111:1611–8. [9] Azar RR, Waters DD, McKay RG, et al. Short- and medium-term outcome differences in women and men after primary percutaneous transluminal mechanical revascularization for acute myocardial infarction. Am J Cardiol 2000;85:675–9. [10] Carrabba N, Santoro GM, Balzi D, et al. AMI-Florence Working Group. In-hospital management and outcome in women with acute myocardial infarction (data from AMI-Florence Registry). Am J Cardiol 2004;94:1118–23. [11] Levey AS, Stevens LA, Schmid CH, et al. CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–12. [12] Barrett BJ, Parfrey PS. Preventing nephropathy induced by contrast medium. N Engl J Med 2006;354:379–86.
[13] Rao AK, Pratt C, Berke A, et al. Thrombolysis in Myocardial Infarction (TIMI) Trial— phase I: hemorrhagic manifestations and changes in plasma fibrinogen and the fibrinolytic system in patients treated with recombinant tissue plasminogen activator and streptokinase. J Am Coll Cardiol 1988;11:1–11. [14] Nutritional anaemias. Report of a WHO scientific group. World Health Organ Tech Rep Ser 1968;405:5–37. [15] 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. [16] Wi J, Ko YG, Kim JS, et al. Impact of contrast-induced acute kidney injury with transient or persistent renal dysfunction on long-term outcomes of patients with acute myocardial infarction undergoing percutaneous coronary intervention. Heart 2011;97:1753–7. [17] Senoo T, Motohiro M, Kamihata H, et al. Contrast-induced nephropathy in patients undergoing emergency percutaneous coronary intervention for acute coronary syndrome. Am J Cardiol 2010;105:624–8. [18] Gill NK, Piccione EA, Vido DA, Clark BA, Shannon RP. Gender as a risk factor for contrast nephropathy: effects of hydration and N-acetylcysteine. Clin Cardiol 2004;27:554–8. [19] Neugarten J, Kasiske B, Silbiger SR, Nyengaard JR. Effects of sex on renal structure. Nephron 2002;90:139–44. [20] Lawesson SS, Alfredsson J, Fredrikson M, Swahn E. Time trends in STEMI-improved treatment and outcome but still a gender gap: a prospective observational cohort study from the SWEDEHEART register. BMJ Open Mar. 27 2012;2:e000726. [21] Sadowski M, Gasior M, Gierlotka M, Janion M, Poloński L. Gender-related differences in mortality after ST-segment elevation myocardial infarction: a large multicenter national registry. Eurointervention 2011;6:1068–72. [22] Lawesson SS, Alfredsson J, Fredrikson M, Swahn E. A gender perspective on shortand long term mortality in ST-elevation myocardial infarction—a report from the SWEDEHEART register. Int J Cardiol 2013;168(2):1041–7.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Female gender and contrast-induced nephropathy in primary percutaneous intervention for ST-segment elevation myocardial infarction Stefano Lucreziotti a,⁎,1, Marco Centola a,1, Diego Salerno-Uriarte a,1, Giorgio Ponticelli a,1, Pier Maria Battezzati b,1, Diego Castini a,1, Carlo Sponzilli a,1, Federico Lombardi c,1 a b c
Cardiologia, Ospedale San Paolo, Dipartimento di Scienze della Salute, University of Milan, Italy Divisione di Medicina VI, Ospedale San Paolo, Dipartimento di Scienze della Salute, University of Milan, Italy U.O.C. di Malattie Cardiovascolari, Fondazione IRCCS, Ospedale Maggiore Policlinico, Dipartimento di Scienze Cliniche e di Comunità, University of Milan, Italy
a r t i c l e
i n f o
Article history: Received 11 February 2013 Received in revised form 10 October 2013 Accepted 14 March 2014 Available online 20 March 2014 Keywords: Contrast-induced nephropathy Coronary angioplasty Female gender Mortality Acute myocardial infarction
a b s t r a c t Background: Patients undergoing primary percutaneous coronary intervention (PCI) are at high risk for contrastinduced nephropathy (CIN), a complication that has been demonstrated to negatively affect outcomes. It has been suggested that, when compared to males, female patients present higher incidence of CIN and higher mortality after primary PCI. However, the specific role of gender in this setting remains ill-defined given its complex interplay with several co-morbidities and clinical characteristics. We investigated the relationship of patients' variables, including gender, with CIN and mortality after primary PCI. Methods: In a single center study in 323 consecutive patients undergoing primary PCI, the development of CIN and mortality during an 18-month median follow-up period was assessed. CIN was defined as an increase in serum creatinine (≥25% or ≥0.5 mg/dl) from baseline occurring at any time during the first 3 post-procedural days. Results: CIN occurred in 23 female and 26 male patients (25.0% vs 11.2%, p = 0.003), while cumulative mortality was 10.6%. Women presented unfavorable basal characteristics and underwent myocardial reperfusion less quickly. At multivariable analysis, reduced left ventricular ejection fraction (LVEF) (odds ratio [OR] 7.32 95% confidence interval [CI]: 2.60–21, p b 0.001) and female gender (OR 2.49 95%CI 1.22–5.07, p = 0.01) predicted CIN, whereas the occurrence of CIN (hazard ratio [HR] 3.65 95%CI 1.55–8.59, p = 0.003) and a Mehran risk score (MRS) ≥6 (HR 1.76 95%CI 1.13–2.74, p = 0.01) independently predicted long-term mortality. Conclusions: After primary PCI, female gender and LVEF are associated with an increased risk of CIN, whereas MRS and development of CIN predict long-term mortality. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Primary percutaneous coronary intervention (PCI) is the best reperfusion strategy in patients presenting with ST-segment elevation myocardial infarction (STEMI) [1,2]. When compared to elective procedures, primary PCI is associated with a higher incidence of contrast-induced nephropathy (CIN), a complication which is associated with increased in-hospital and long-term morbidity and mortality [3].
⁎ Corresponding author at: Cardiologia, Ospedale San Paolo, Via A. di Rudinì 8, 20142 Milan, Italy. Tel./fax: +39 02 50323145. E-mail address: [email protected] (S. Lucreziotti). 1 This author takes responsibility of all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2014.03.087 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.
It would be therefore useful and desirable to promptly detect patients at high risk for CIN because they could be carefully monitored and treated with adequate prophylactic strategy. Patients prone to develop CIN are those with older age, hemodynamic instability, long symptom to reperfusion time, and anterior infarction [3]. Although it has been suggested that female gender exposes patients to higher risk for CIN, no conclusive evidence is available, particularly in the setting of primary PCI [3–7]. Uncertainty also exists regarding gender differences in survival after primary PCI: women have a worse short- and long-term prognosis than men, even though this difference seems to be largely influenced by comorbidities and clinical variables, which characterize female gender [8–10]. The aim of the present study was to investigate the correlation between female gender and CIN in a cohort of STEMI patients undergoing primary PCI. The role of female gender and CIN in predicting mortality was also evaluated.
38
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
2. Methods 2.1. Patients Data on STEMI patients admitted from January 2009 to December 2011 were prospectively entered and retrospectively analysed. We enrolled all consecutive patients who presented to the emergency department of our hospital within 12 h (24 h in the presence of cardiogenic shock or clinical/electrocardiographic evidence of ongoing ischaemia) from the onset of symptoms with a ST-segment elevation electrocardiographic pattern in two or more contiguous leads, or left bundle branch block. Exclusion criteria were: cardiac arrest requiring a prolonged resuscitation (N30 min), ongoing dialytic treatment, and malignancy with a life expectancy of less than 6 months. All patients were followed up by means of a hospital visit at 6 months and every subsequent year or by telephone interviews. This investigation was designed as an observational study and it conforms to the principles outlined in the Declaration of Helsinki; the Research Ethics Board of the San Paolo Hospital and the University of Milan, Italy approved the study protocol. An informed consent was obtained from all patients. 2.2. Coronary angioplasty and medical therapy Before undergoing coronary angiography, all patients underwent a rapid clinical assessment with blood sampling and 12-lead ECG recording. Medical therapy and modality of revascularization were left at the discretion of the coronary care and interventional cardiologists. Prophylaxis for CIN was not routinely administered. All patients received iso-osmolar, nonionic contrast agent iodixanol. Primary PCI was performed by an interventional team on a 24-hour basis with a standard technique via femoral or radial approach. Periprocedural antithrombotic therapy was prescribed according to the current guidelines [1]. Briefly, a loading dose of aspirin (300 to 500 mg) and clopidogrel (300 or 600 mg) or prasugrel (60 mg) was administered before the procedure, while anticoagulant therapy was started after arterial sheath insertion. Heparin was administered at the doses of 100 U/kg or 70 U/kg, if an additional therapy with abciximab was planned. Additional heparin boluses were eventually added in order to maintain the activated clotting time N300 s, or between 200 and 250 if abciximab was used. Intravenous bivalirudin (0.75 mg/kg bolus and 1.75 mg/kg/h infusion for the duration of the procedure) was an alternative anticoagulant therapy. An early removal of arterial sheath was encouraged. In patients with multivessel disease, only the infarctrelated artery was treated during the acute phase. Delayed percutaneous or surgical revascularization was eventually performed in a later time (after at least 4 days, if the patients were clinically stable) in case of critical stenosis of one or more coronary arteries supplying a significant amount of jeopardized myocardium. 2.3. Other diagnostic techniques and definitions Echocardiographic assessment of the left ventricular ejection fraction (LVEF) was performed within 24 h after primary PCI. Values b40% were considered to reflect a reduced LVEF. Serum creatinine was measured at hospital admission and every day for the following 3 days. Serum creatinine levels ≥1.2 mg/dl were considered elevated. Estimated glomerular filtration rate (eGFR) was calculated according to the CKD-EPI equation [11]. CIN was defined as an increase ≥25% or ≥0.5 mg/dl (44 mmol/l) in serum creatinine from the baseline values within the first 72 h after the index procedure [12]. Major bleeding was defined according to the ‘Thrombolysis in Myocardial Infarction’ (TIMI) criteria [13]. A hemoglobin value b13 g/dl for men and b12 g/dl for women defined anemia [14]. Time to reperfusion was calculated as an interval between the symptom onset and the first balloon inflation. The risk for CIN was calculated in relation to the Mehran risk score (MRS) [15]. According to value distribution, patients were classified as low or medium-to-high risk when MRS was b or ≥6.
Death from any cause was the other major clinical outcome of the study. To summarize time to death data, death rates from any cause were expressed as the number of patients dying per 100 person-months, with death rates ratios and 95% CI reported to contrast risk groups. Cumulative proportions of patients dying were estimated by means of the Kaplan and Meier approach using the date of presentation to the Emergency Department as the starting point. Univariable analyses of risk factors for death were performed using the log-rank test. The proportional hazard Cox model was used to assess within the framework of multivariable analysis the effects of risk factors associated (p b 0.10) with death at univariable analysis while simultaneously adjusting for the same set of predefined variables used in the models predicting development of CIN. Likelihood ratio tests were used to test the joint effect of adding, or omitting, specific groups of variables on the predictive value of a Cox model. To avoid problems inherent in the categorization of continuous variables, noncategorical variables were preferentially introduced in multivariable models as continuous variables, unless categorization was strongly suggested by widespread prior use. For example, the MRS was introduced as categorical variable with score values grouped in 4 classes (MRS b6, 6–10, 11–16, or N16). In order to facilitate interpretation and visualization of results, in some instances analyses were also done arranging patients in two groups, i.e., those with low (b6) or moderate-to-high (≥6) MRS values. Statistical comparisons were performed using Stata Statistical Software (version 11, Stata Corporation, College Station, TX). All analyses were two-sided.
3. Results A total of 332 consecutive patients were considered. Nine patients were excluded from the study for the following reasons: dialysis (2 patients), coronary anatomy not suitable for PCI (4 patients), and death within the first 24 h after PCI (3 patients). Hence, 323 patients were included in the analysis. Follow-up was completed in 322 (99.7%) patients. Median follow-up period was 18 months (interquartile range 9.4–29.7 months) with an overall duration of 522 patient-years. 3.1. Patient characteristics and incidence of CIN Major clinical and procedural characteristics of the whole patients' group and of those who did or did not develop CIN are reported in Table 1. Seventy-nine (24.5%) patients were older than 75 years, and 55 (17%) had diabetes. Only a relatively small amount (9.3%) of subjects presented a reduced LVEF. Most patients were in Killip class 1 (87%) and in the low MRS class (64.3%). The left anterior coronary artery was the infarct-related vessel in 150 patients (46.4%). A multivessel disease was detected in 193 (59.8%) patients. The rate of major in-hospital bleeding was 4.7%. The mean value of creatinine was 1.0 ± 0.5 mg/dl. Serum creatinine levels were elevated in 56 (17.3%) patients. CIN was detected in 49 (15.2%) subjects, with a significantly (p = 0.003) higher incidence in women (25.0%) than in men (11.3%). 3.2. Predictors of CIN
2.4. Statistical analysis Continuous variables are summarized as mean ± SD values. The significance of differences between groups was assessed by the Mann–Whitney test, or by the Kruskal–Wallis test when more than two groups were concerned. Differences in proportions were tested using Fisher's exact test or chi-square statistics. The major outcomes of this study were development of CIN and death from any cause. Risk ratios (RR) and their 95% confidence intervals (CI) along with p values were used to describe the risk of developing CIN relative to a reference category. Mantel–Haenszel tests for linear trends were used to assess the presence of trends across different patients' categories or different levels of exposure. Logistic regression analysis was used to adjust risks, expressed as odd ratios (OR), while simultaneously adjusting for the effect of other variables under study. As stipulated in the study protocol, variables associated (p b 0.10) with the two major study outcome variables associated (p b 0.10) with development of CIN at univariable analysis were tested in multivariable models adjusted for the following variables: systolic blood pressure, hypertension, family history of cardiovascular events, anterior myocardial infarction, time to reperfusion, and multivessel coronary disease. Such variables were introduced regardless of their significance level in view of their potential role as confounding factors. Because MRS is a validated risk score for CIN and clinical outcome after PCI [7, 15], we did not include in multivariable models those variables concurring to the calculation of MRS and those directly associated with them, i.e., we did not include age, serum hemoglobin, diabetes, contrast media volume, serum creatinine and eGFR to avoid overfitting model or multicollinearity problems.
As shown in Table 1, CIN was more frequent in women and in patients with advanced age, reduced LVEF, impaired renal function and higher MRS. Mean MRS was 6.5 ± 4.8 in the patients who developed CIN compared with 5.2 ± 4.6 (p = 0.07) in those who did not develop the complication. The difference in BMI was of borderline significance, and largely depended on the different distribution of CIN between genders. When the distribution of BMI values into gender-specific quintiles was compared between patients according to development of CIN, such difference fell well below the significance level (p = 0.81). A similar result was found when BMI values were compared after adjustment for gender by logistic regression analysis (p = 0.37). The amount of contrast medium, the frequency of major bleeding, the basal glucose level, the electrocardiographic location of myocardial infarction and the Killip class were not significantly different between patients who developed or did not develop CIN. There were some relevant gender differences in basal and procedural risk profile (Table 2): women were older, and presented a higher MRS and a lower renal function. Mean MRS was 7.3 ± 5.0 in female patients compared with 4.7 ± 4.3
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
39
Table 1 Clinical and procedural characteristics of patients developing CIN. Variable
All patients (n = 323)
Patients with CIN (n = 49)
Patients without CIN (n = 274)
p value
Age, years Age ≥75 years Women Active smoking Diabetes mellitus Family history of cardiovascular disease Hypercholesterolemia Systolic blood pressure, mm Hg Hypertension BMI, kg/m2 LVEF b40% Anterior myocardial infarction Killip class 2–4 Multivessel coronary disease In-hospital major bleeding Time to reperfusion, min Contrast volume, ml Contrast volume N300 ml Hemoglobin, g/dl Glucose, mg/dl Creatinine, mg/dl eGFR, ml/min/1.73 m2 eGFR b60 ml/min/1.73 m2 MRS class b6 6–10 11–16 N16 MRS ≥6
65.3 ± 13.0 79 (24.5) 92 (28.5) 138 (42.7) 55 (17.0) 55 (17.0) 107 (33.1) 144 ± 27 169 (52.3) 26.1 ± 3.6 30 (9.3) 146 (45.2) 42 (13.0) 193 (59.8) 15 (4.7) 302 ± 320 259 ± 101 70 (21.7) 14.3 ± 1.8 159 ± 69 1.0 ± 0.5 76.7 ± 23.0 77 (23.8)
69.3 ± 14.7 18 (36.7) 23 (46.9) 20 (40.8) 10 (20.4) 6 (12.2) 12 (24.5) 148 ± 24 24 (49.0) 25.4 ± 3.8 13 (26.5) 24 (49.0) 9 (18.4) 27 (55.1) 5 (10.2) 341 ± 324 245 ± 85 7 (14.3) 14.0 ± 2.0 175 ± 85 1.2 ± 1.0 72.1 ± 31.6 20 (40.8)
64.6 ± 12.6 61 (22.3) 69 (25.2) 118 (43.1) 45 (16.4) 49 (17.9) 95 (34.7) 143 ± 27 145 (52.9) 26.2 ± 3.6 17 (6.2) 122 (44.5) 33 (12.0) 166 (60.6) 10 (3.7) 295 ± 320 262 ± 104 63 (23.0) 14.4 ± 1.8 156 ± 65 1.0 ± 0.4 77.5 ± 21.1 57 (20.8)
0.03 0.05 0.003 0.88 0.54 0.41 0.19 0.53 0.64 0.09 b0.001 0.64 0.25 0.53 0.06 0.12 0.4 0.19 0.11 0.17 0.28 0.58 0.006
207 (64.3) 69 (21.4) 36 (11.2) 10 (3.1) 115 (35.7)
23 (47.9) 14 (29.2) 9 (18.8) 2 (4.2) 25 (52.1)
184 (67.2) 55 (20.1) 27 (9.9) 8 (2.9) 90 (32.9)
0.07
0.01
Values are n (%) or mean ± SD, unless otherwise specified. CIN; contrast-induced nephropathy, BMI; body mass index, LVEF; left ventricular ejection fraction, eGFR; estimated glomerular filtration rate, MRS; Mehran risk score.
(p b 0.001) in males. In addition, women were treated with a considerable time-delay compared with males. On the other hand, males were more frequently, smokers. We have found a significant (p b 0.001) trend toward increasing risk of CIN across different gender and LVEF categories, with a strong interaction between these two variables (Fig. 1). At multivariable analysis (Table 3), female gender (OR 2.49 95%CI 1.22–5.07, p = 0.01) and reduced LVEF (OR 7.32 95%CI 2.60– 21, p b 0.001) were accompanied with significantly increased risk of CIN. In contrast, we did not observe any statistical association between MRS, in-hospital bleeding and CIN.
3.3. In-hospital and long-term mortality During hospital stay, 17 (5.3%) deaths occurred in the study population. In-hospital mortality was higher in patients developing CIN than in patients without this complication (20.4% vs 2.6%, RR 7.99 95%CI 3.19–20, p b 0.001), in women (9.8% vs 3.5%, RR 2.82 95%CI 1.12–7.10, p = 0.02), and in subjects with reduced LVEF (36.7% vs 2.0%, RR 17 95%CI 7.12–45, p b 0.001). When MRS was categorized into two classes, mortality was 11.3% in patients showing moderate-to-high values and 1.9% in those showing an MRS b 6 (RR 5.85 95%CI 1.95–18, p b 0.001).
Table 2 Clinical and procedural characteristics according to sex. Variable
Women (n = 92)
Men (n = 231)
p value
Age, years Active smoking Diabetes mellitus Family history of cardiovascular disease Hypercholesterolemia Hypertension LVEF b40% Killip class 2–4 Anterior myocardial infarction Time to reperfusion, min Contrast volume, ml MRS ≥6 Hemoglobin, g/dl Creatinine, mg/dl eGFR, ml/min/1.73 m2 eGFR b60 ml/min/1.73 m2 Glucose, mg/dl
72.0 ± 13.5 25 (27.2) 19 (20.7) 15 (16.3) 33 (35.9) 46 (50.0) 13 (14.1) 16 (17.4) 40 (43.5) 363 ± 357 253 ± 86 51 (56.0) 13.1 ± 1.5 1.0 ± 0.8 70.0 ± 26.1 36 (39.1) 161.1 ± 69.4
62.7 ± 11.8 113 (48.9) 36 (15.6) 40 (17.3) 74 (32.0) 123 (53.3) 17 (7.4) 26 (11.3) 106 (45.9) 278 ± 302 262 ± 107 64 (27.7) 14.8 ± 1.7 1.1 ± 0.4 79.4 ± 21.1 41 (17.7) 159.0 ± 69.0
b0.001 b0.001 0.33 0.87 0.52 0.62 0.09 0.15 0.71 0.002 0.81 b0.001 b0.001 b0.001 0.003 b0.001 0.68
Values are n (%) or mean ± SD, unless otherwise specified. BMI; body mass index, LVEF; left ventricular ejection fraction, MRS; Mehran risk score, eGFR; estimated glomerular filtration rate.
40
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42 Table 4 Factors associated with long-term-mortality at univariable analysis.
Fig. 1. Relative risk of CIN according to gender and LVEF. Male patients with LVEF ≥40% are the reference group, i.e., risk ratio is 1. Vertical bars denote 95% CI for the risk ratios; p values refer to the comparison with reference group. Mantel–Haenszel test for trend across the four categories: p b 0.001. CIN; contrast-induced nephropathy, LVEF; left ventricular ejection fraction, CI; confidence interval.
During the 18 month follow-up, 33 (10.6%) patients died. Clinical variables potentially associated with mortality were evaluated with univariable and multivariable analysis (Tables 4, 5). The multivariable analysis showed that only a moderate-to-high MRS and the occurrence of CIN were significantly associated with the risk of death, whereas the relationships between in-hospital major bleeding, reduced LVEF, female gender and long-term mortality were not statistically significant. Interestingly, there was a significant trend (p b 0.001) toward increasing mortality across the groups defined by moderate-to-high MRS and development of CIN (Fig. 2). 4. Discussion The main finding of the present study was that female gender was an independent predictor of CIN in STEMI patients treated with primary PCI. CIN development and high MRS were associated with increased mortality independently of gender. Previous studies have shown that the risk for CIN was higher in women undergoing elective or urgent PCI [5,6,16,17], but a correlation between female gender and CIN has never been reported in a selected population of STEMI patients treated by primary PCI [3,7]. A reduced glomerular volume in women compared to their male counterpart has been proposed as a possible mechanism to explain a greater susceptibility to acute renal failure [18]. It cannot be ruled out, however, that the association between female gender and CIN rather to be causal might be secondary to the higher prevalence of comorbidities that characterize female gender or, finally, to older age in our female population.
Table 3 Predictive factors for CIN at multivariable analysis. Variable
Odds ratio
95% CI
p value
LVEF b40% Female gender MRS class In-hospital major bleeding
7.32 2.49 1.02 2.61
2.60–21 1.22–5.07 0.63–1.64 0.70–9.77
b0.001 0.01 0.95 0.15
CIN; contrast induced nephropathy, OR; odds ratio, CI; confidence interval, LVEF; left ventricular ejection fraction, MRS; Mehran risk score.
Variable
Death rate
Death rate ratio
95% CI
p value
Age ≥75 years CIN Female gender Diabetes mellitus Hypertension LVEF b40% Anterior myocardial infarction Killip class 2–4 Multivessel coronary disease In-hospital major bleeding Time to reperfusiona Contrast volume N300 ml Anemia High Glucose levelb eGFR b60 ml/min/1.73 m2 MRS class b6 6–10 1–16 ≥16
1.93 2.05 1.06 0.75 0.60 3.79 0.62 1.94 0.62 2.70 0.48 0.55 1.13 0.67 2.24
2.82 6.03 3.07 1.55 1.37 11.23 1.36 5.41 1.62 5.86 0.88 1.05 3.00 1.37 12.93
4.49–23 2.80–13 1.46–6.50 0.61–3.56 0.65–3.00 5.13–23 0.65–2.89 2.47–11 0.73–3.97 1.77–15 0.30–2.19 0.41–2.41 1.39–6.31 0.56–3.06 5.80–32
b0.001 b0.001 0.001 0.26 0.31 b0.001 0.43 b0.001 0.13 0.001 0.79 0.75 0.002 0.27 b0.001
0.16 0.93 1.92 4.08
1 5.88 12.10 25.77
2.08–18 4.16–37 6.45–94
b0.001
CI; confidence interval, CIN; contrast-induced nephropathy, BMI; body mass index, LVEF; left ventricular ejection fraction, eGFR; estimated glomerular filtration rate, MRS; Mehran risk score. a Values in the higher quintile of the distribution of time to reperfusion. Cut-off level is 375 min. b Values in the higher quintile of the distribution of serum glucose level. Cut-off level is 190 mg/dl.
The new finding of an association between female gender and CIN is of relevant clinical and physiopathological significance but must be interpreted by considering some methodological features that characterize our study in comparison with previous ones. First, unlike previous STEMI studies [3,7], periprocedural hydration was not routinely administered in our study. Our population was therefore exposed to a higher risk of CIN due to the lack of prophylactic hydration that appears be more effective in women than in men [19]. Second, we determined CIN according to the recently validated CKD-EPI equation [11], at variance with previous studies where a different formula was used [3,7]. Regarding this methodological aspect, it has been reported that the CKD-EPI may provide a better evaluation of renal function than the MDRD formula, especially when measuring eGFR in a population that consists of patients with a preserved renal function [11]. Third, we included patients with cardiogenic shock who have been previously excluded in earlier studies [7]. Fourth, we defined CIN in relation to either absolute or percent increase in serum creatinine, whereas, for example, Marenzi et al. only considered absolute increases of this parameter [3]. Nevertheless, the incidence of CIN observed in our study was similar to that previously reported thus making unlikely that these methodological aspects might have played a major role. Finally, we also observed that a reduced LVEF was associated with an increased risk of CIN, whereas other traditional risk factors including the amount of injected contrast medium did not predict CIN development. This negative finding, which is in disagreement with previous reports [3,16], may Table 5 Predictors of long-term mortality at multivariable analysis. Variable
Hazard ratio
95% CI
p value
MRS class a CIN LVEF b40% In-hospital major bleeding Female gender
1.76 3.65 2.41 1.52 1.71
1.13–2.74 1.55–8.59 0.87–6.71 0.46–4.99 0.82–3.59
0.012 b0.003 0.09 0.49 0.15
CI; confidence interval, MRS; Mehran risk score, CIN; contrast-induced nephropathy, LVEF; left ventricular ejection fraction. a MRS was categorized in four classes. The reported hazard ratio and 95% CI are interpreted as the estimated hazard associated to a one-unit increase in MRS class.
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
41
Fig. 2. Kaplan–Meier estimates of mortality according to MRS and development of CIN. Patients are categorized based on the presence of a low (values b6) or a moderate-to-high (values ≥6) MRS and according to whether they did or did not develop CIN. MRS; Mehran risk score, CIN; contrast-induced nephropathy.
have been explained by the fact that in our center a selective strategy of revascularization limited to the infarct-related artery was adopted in all STEMI patients. Hence, the relatively low and less variable amount of injected contrast medium could account for the lack of association between this parameter and risk of CIN. Considerable debate exists whether gender has an independent prognostic role after STEMI. The present study shows that, after adjustment for other variables including CIN and MRS, long-term mortality after primary PCI was not higher in women than in men. These findings are consistent with the results of previous trials and registries [8–10, 20–22]. For example, a large Swedish registry has recently demonstrated a relevant gender gap regarding treatment appropriateness and mortality in patients with STEMI [20]; nevertheless, one year mortality was not higher in women when adherence to the treatment guidelines was considered in the multivariable analysis. Similar results were also observed in a large multicenter registry that also reported the presence of a significant difference in the rate of reperfusion treatment [21]. A lack of gender effect on long-term mortality was also found in two single center registries and in one multicenter, randomized, controlled trial, which, as in our study, enrolled only STEMI patients treated with primary PCI [8–10]. All these data suggest that women still receive suboptimal treatment and are more exposed to hard events after hospital discharge in the era of widespread use of reperfusion therapy and potent antithrombotic therapies. In our study, for example, women undergoing primary PCI presented a less favorable basal clinical profile and a longer reperfusion time. In addition, we observed that CIN development accounted for the higher in-hospital and overall mortality, a finding that further confirms the negative prognostic impact of this complication. The capability of MRS to predict mortality as well as major adverse cardiovascular and cerebrovascular events in STEMI patients treated with primary PCI has been previously reported [7]. We also found that MRS was clinically useful in predicting long-term mortality in this clinical context, thus confirming the predicting value for hard cardiovascular events of what has been considered a simple “renal” risk score. Quite surprisingly, we were unable to detect an association between MRS and CIN that has been previously proposed and validated in a large series of patients who underwent PCI, excluding those with an acute
myocardial infarction [15]. The different patients' subset, the inclusion of gender among the variables tested by multivariable analysis, the relative small sample size and the inhomogeneous distribution of patients in the four classes of MRS (Table 1) may account for this negative finding. Finally, it must be pointed out that we did not assess the independent prognostic value of those clinical or laboratory variables (for example, age and creatinine plasma levels) included in MRS.
5. Limitations This study was of smaller size compared with some of the available ones: this may have limited the statistical power to detect possible additional associations. It should be noted, however, that we enrolled an unselected population of all comers for STEMI who have been uniformly treated with primary PCI, as in the everyday clinical practice. This is a single-center study but with the advantage of a more homogeneous patients' management. We can't exclude that in some patients factors other than contrast medium including renal embolisms or dehydration might have contributed to the raise in plasma creatinine levels. Due to the limited number of deaths, only total mortality was considered. Finally, not all variables potentially affecting prognosis were evaluated in the present study.
6. Conclusions In patients undergoing primary PCI, female gender and reduced LVEF are associated with CIN risk. Both CIN and MRS, but not female gender, are independent predictors of long-term mortality at multivariable analysis.
References [1] Steg PG, James SK, Atar D, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation: the Task Force on the management of ST-segment elevation acute myocardial infarction of the European Society of Cardiology (ESC). Eur Heart J 2012;33:2569–619.
42
S. Lucreziotti et al. / International Journal of Cardiology 174 (2014) 37–42
[2] Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:13–20. [3] Marenzi G, Lauri G, Assanelli E, et al. Contrast-induced nephropathy in patients undergoing primary angioplasty for acute myocardial infarction. J Am Coll Cardiol 2004;44:1780–5. [4] Mueller C, Buerkle G, Perruchoud AP, Buettner HJ. Female sex and risk of contrast nephropathy after percutaneous coronary intervention. Can J Cardiol 2004;20:505–9. [5] Sidhu RB, Brown JR, Robb JF, et al. Interaction of gender and age on post cardiac catheterization contrast-induced acute kidney injury. Am J Cardiol 2008;102:1482–6. [6] Thompson CA, Kaplan AV, Friedman BJ, et al. Gender-based differences of percutaneous coronary intervention in the drug-eluting stent era. Catheter Cardiovasc Interv 2006;67:25–31. [7] Sgura FA, Bertelli L, Monopoli D, et al. Mehran contrast-induced nephropathy risk score predicts short- and long-term clinical outcomes in patients with STelevation–myocardial infarction. Circ Cardiovasc Interv 2010;3:491–8. [8] Lansky AJ, Pietras C, Costa RA, et al. Gender differences in outcomes after primary angioplasty versus primary stenting with and without abciximab for acute myocardial infarction: results of the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial. Circulation 2005;111:1611–8. [9] Azar RR, Waters DD, McKay RG, et al. Short- and medium-term outcome differences in women and men after primary percutaneous transluminal mechanical revascularization for acute myocardial infarction. Am J Cardiol 2000;85:675–9. [10] Carrabba N, Santoro GM, Balzi D, et al. AMI-Florence Working Group. In-hospital management and outcome in women with acute myocardial infarction (data from AMI-Florence Registry). Am J Cardiol 2004;94:1118–23. [11] Levey AS, Stevens LA, Schmid CH, et al. CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–12. [12] Barrett BJ, Parfrey PS. Preventing nephropathy induced by contrast medium. N Engl J Med 2006;354:379–86.
[13] Rao AK, Pratt C, Berke A, et al. Thrombolysis in Myocardial Infarction (TIMI) Trial— phase I: hemorrhagic manifestations and changes in plasma fibrinogen and the fibrinolytic system in patients treated with recombinant tissue plasminogen activator and streptokinase. J Am Coll Cardiol 1988;11:1–11. [14] Nutritional anaemias. Report of a WHO scientific group. World Health Organ Tech Rep Ser 1968;405:5–37. [15] 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. [16] Wi J, Ko YG, Kim JS, et al. Impact of contrast-induced acute kidney injury with transient or persistent renal dysfunction on long-term outcomes of patients with acute myocardial infarction undergoing percutaneous coronary intervention. Heart 2011;97:1753–7. [17] Senoo T, Motohiro M, Kamihata H, et al. Contrast-induced nephropathy in patients undergoing emergency percutaneous coronary intervention for acute coronary syndrome. Am J Cardiol 2010;105:624–8. [18] Gill NK, Piccione EA, Vido DA, Clark BA, Shannon RP. Gender as a risk factor for contrast nephropathy: effects of hydration and N-acetylcysteine. Clin Cardiol 2004;27:554–8. [19] Neugarten J, Kasiske B, Silbiger SR, Nyengaard JR. Effects of sex on renal structure. Nephron 2002;90:139–44. [20] Lawesson SS, Alfredsson J, Fredrikson M, Swahn E. Time trends in STEMI-improved treatment and outcome but still a gender gap: a prospective observational cohort study from the SWEDEHEART register. BMJ Open Mar. 27 2012;2:e000726. [21] Sadowski M, Gasior M, Gierlotka M, Janion M, Poloński L. Gender-related differences in mortality after ST-segment elevation myocardial infarction: a large multicenter national registry. Eurointervention 2011;6:1068–72. [22] Lawesson SS, Alfredsson J, Fredrikson M, Swahn E. A gender perspective on shortand long term mortality in ST-elevation myocardial infarction—a report from the SWEDEHEART register. Int J Cardiol 2013;168(2):1041–7.
