Author + information
- Received October 29, 2012
- Revision received February 25, 2013
- Accepted March 13, 2013
- Published online October 1, 2013.
- Yohei Ohno, MD∗,
- Yuichiro Maekawa, MD∗,
- Hiroaki Miyata, PhD†,
- Soushin Inoue, MD‡,
- Shiro Ishikawa, MD§,
- Koichiro Sueyoshi, MD‖,
- Shigetaka Noma, MD¶,
- Akio Kawamura, MD∗,
- Shun Kohsaka, MD∗∗ ( and )
- Keiichi Fukuda, MD, PhD∗
- ∗Department of Cardiology, Keio University School of Medicine, Tokyo, Japan
- †Department of Healthcare Quality Assessment, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
- ‡Department of Cardiology, Hino Municipal Hospital, Tokyo, Japan
- §Department of Cardiology, Saitama Municipal Hospital, Saitama, Japan
- ‖Department of Cardiology, Kawasaki Municipal Hospital, Kanagawa, Japan
- ¶Department of Cardiology, Saiseikai Utsunomiya Hospital, Tochigi, Japan
- ↵∗Reprint requests and correspondence:
Dr. Shun Kohsaka, Department of Cardiology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku, Tokyo 160-8582, Japan.
Objectives This study sought to evaluate the association between contrast-induced acute kidney injury (CI-AKI) after percutaneous coronary intervention and severity of bleeding estimated from periprocedural hemoglobin (Hb) measurement.
Background The relationship between CI-AKI and bleeding in contemporary practice remains controversial.
Methods In a retrospective analysis of the prospectively maintained Japan Cardiovascular Database-Keio Interhospital Cardiovascular Studies (JCD-KICS) multicenter registry, we divided 2,646 consecutive patients into 5 groups according to the change of Hb level after compared with before percutaneous coronary intervention: patients without a decrease in Hb level (group A) and patients with a decreased Hb level: <1 g/dl (group B); 1 to <2 g/dl (group C); 2 to <3g/dl (group D); and >3 g/dl (group E). CI-AKI was defined as an increase in serum creatinine level ≥0.5 mg/dl or ≥25% above baseline values at 48 h after administration of contrast media. Procedure and outcome variables were compared.
Results The mean patient age was 67 ± 11 years. Of the 2,646 patients, CI-AKI developed in 315 (11.9%). The CI-AKI incidence was 6.2%, 7.5%, 10.7%, 17.0%, and 26.2%, in groups A through E, respectively (p < 0.01), whereas the incidence of major bleeding was 0.7%, 1.3%, 2.0%, 4.1%, and 28.3%, respectively (p < 0.01). CI-AKI was associated with higher rates of mortality (5.4% vs. 0.6%, p < 0.01) and of composite of heart failure, cardiogenic shock, and death (16.5% vs. 2.8%, p < 0.01).
Conclusions Periprocedural bleeding was significantly associated with CI-AKI, with CI-AKI incidence correlating with bleeding severity.
Contrast-induced acute kidney injury (CI-AKI) is one of the most common major adverse events after percutaneous coronary intervention (PCI) and is associated with increased morbidity, short- and long-term mortality, and health-care costs (1,2). Risk factors for development include volume and type of contrast agent, and patient-related factors such as chronic kidney disease (CKD), diabetes mellitus, congestive heart failure, advanced age, anemia, sex, hemodynamic instability, and reduced effective circulating volume (3,4). In addition, the cumulative effect of these parameters on renal function has been assessed with risk score models in a number of studies (5,6). Thus far, prophylactic strategies for CI-AKI are mainly related to preprocedural measurements such as use of hydration, bicarbonate, and N-acetylcysteine, which remain controversial (7,8). Definition of other risk factors for CI-AKI, particularly procedural ones may help broaden prophylactic strategies.
Recent studies have documented an association between bleeding and mortality (short and long term) in patients with acute coronary syndromes (ACS) as well as in those undergoing PCI (9,10). Bleeding complications are not uncommon, but may be underrecognized, and prompt, effective identification and management after PCI remains a major challenge. This may have a deleterious impact on prognosis, outcomes, and costs. Anemia is known to be an independent predictor of CI-AKI (4), being one of the main variables of the Mehran contrast-induced nephropathy risk score (5). However, few previous studies or risk score models have accounted for periprocedural bleeding as a risk factor for the development of CI-AKI. If periprocedural bleeding is strongly associated with CI-AKI incidence, there may be a potential to reduce the latter by recognizing and actively managing the extent of periprocedural bleeding. The purpose of this study was to determine CI-AKI incidence after PCI and to evaluate the impact of bleeding on CI-AKI development.
The JCD-KICS (Japan Cardiovascular Database-Keio Interhospital Cardiovascular Studies) is a large, ongoing, prospective multicenter cohort registry designed to collect clinical data on PCI patients. Patients were enrolled at the time of their first coronary event, and all consecutive PCI procedures during the study period were registered, including failure cases. The JCD-KICS began enrolling patients in September 2008, and participating hospitals collate patient data into an Internet-based database system. The assigned clinical coordinators review the system periodically to ensure that the reported data are accurately and consistently documented. The majority of clinical variables in JCD-KICS were defined according to the National Cardiovascular Data Registry, which was sponsored by the American College of Cardiology to conduct comparative research to determine the factors that lead to disparities in PCI management.
Before the launch of the JCD-KICS, information on the objectives of the present study, its social significance, and an abstract were provided for clinical trial registration with the University Hospital Medical Information Network (UMIN000004736). The study protocol was approved by the institutional review board at each site.
As of December 31, 2011, 12 hospitals had contributed data on 4,117 consecutive PCI procedures. An interventional team performed PCIs according to standard clinical practice via the femoral or radial approach. Supportive pharmacological therapies, mechanical support, contrast medium dose (nonionic low osmolar), and angioplasty technique were left to the discretion of the operator, according to each institution's clinical protocols and international guidelines. Patients with estimated glomerular ﬁltration rate (eGFR) <60 ml/min received intravenous hydration; however, no patients received bicarbonate or N-acetylcysteine in this study. Nephrotoxic medications such as biguanide or nonsteroidal anti-inflammatory drugs were stopped before elective cases and after admission in emergent cases. For the present analysis, 1,471 patients were excluded because of missing initial or peak serum creatinine (Cr) and/or hemoglobin (Hb) (n = 974) levels or incomplete data on sex, age, or death (n = 216). Furthermore, patients with cardiopulmonary arrest and/or cardiogenic shock (n = 169), those on dialysis (n = 40), and those with Cr level ≥8.0 mg/dl or Hb <7 g/dl (n = 72) were also excluded (Fig. 1). We divided the remaining 2,646 consecutive patients into 5 groups according to the change in Hb after versus before PCI. Patients without a decrease in Hb (group A) and patients with decreased Hb level: <1 g/dl (group B), 1 to <2 g/dl (group C), 2 to <3 g/dl (group D), and >3 g/dl (group E).
Patient demographic data included risk factors for coronary artery disease, presenting condition at the time of PCI (elective or urgently/emergently for ACS), and procedural data.
The recommended antiplatelet regimen was long-term 81 to 162 mg aspirin daily and a thienopyridine (200 mg ticlopidine or 75 mg clopidogrel daily) for at least 1 year after drug-eluting or bare metal stent implantation for patients presenting with ACS. For patients who received a bare metal stent for non-ACS, a thienopyridine was given for a minimum of 1 month and ideally up to 12 months. During PCI, unfractionated heparin was used for anticoagulation to achieve and maintain an activated clotting time of 250 to 300 s.
We investigated all-cause death as well as the composite of heart failure, cardiogenic shock, and death. In addition, we investigated the rates of complications such as periprocedural myocardial infarction, heart failure, cardiogenic shock, stroke, tamponade, new requirement for hemodialysis, need for blood transfusion, and major bleeding.
CI-AKI was defined according to the established definitions in the literature as an increase in serum Cr ≥0.5 mg/dl or ≥25% above baseline values at 48 h after administration of contrast media when no other major kidney insult was identified. If >1 post-PCI Cr level was measured, the highest value was used for CI-AKI calculation. Major bleeding was deﬁned as: 1) bleeding requiring a blood transfusion; 2) a decrease in Hb ≥3.0 g/dl from any location, including percutaneous entry site, retroperitoneal, gastrointestinal, genitourinary, and other/unknown location; and 3) procedural intervention/surgery at the bleeding site to reverse/stop or correct the bleeding (such as surgical closures/exploration of the arteriotomy site, balloon angioplasty to seal an arterial tear, or endoscopy with cautery of a gastrointestinal bleed). The latter definition is equivalent to Bleeding Academic Research Consortium type 3 bleeding (11). Minor bleeding was defined as all other clinically significant bleeding not meeting the definition for major bleeding which corresponds to Bleeding Academic Research Consortium type 2 bleeding.
CKD was deﬁned as a baseline eGFR <60 ml/min/1.73 m2. The eGFR was calculated using serum Cr levels and the Modification of Diet in Renal Disease study equation (12).
Baseline characteristics were compared between the patients with CI-AKI and those without CI-AKI. Continuous variables are expressed as mean ± SD unless otherwise specified. Categorical variables are expressed as absolute values and percentages. The independent-samples t test and analysis of variance (ANOVA) were used for comparison of means of continuous variables. The Pearson chi-square test was used for comparison of categorical variables. Multivariable logistic regression analysis was used to assess variables independently associated with CI-AKI. All p values are 2-sided. Results were considered to be statistically significant at a p value <0.05. All statistical analyses were performed with SPSS version 19.0.0 (SPSS Inc., Chicago, Illinois).
Of the 2,646 patients in the study population, 315 (12%) experienced CI-AKI after PCI. Baseline demographic, clinical, and angiographic characteristics, as well as main procedural data at the time of primary PCI are listed in Table 1. In the overall population, the mean age was 67.6 years, and 80% were males. The mean baseline serum Cr level was 0.9 ± 0.4 mg/dl. There were significant clinical and procedural differences between subjects with and without CI-AKI. As seen in Table 1, the mean baseline serum Cr level was higher with CKD observed more frequently in the CI-AKI group. Patients with CI-AKI also had lower left ventricular ejection fraction and more often had a history of heart failure than patients without CI-AKI. Patients with CI-AKI more frequently underwent urgent/emergent PCI and had more complex coronary disease (3-vessel coronary artery disease, intra-aortic balloon pump use during PCI, and type C lesions). Finally, total fluoroscopy time was longer in the CI-AKI group.
Variables independently associated with CI-AKI after PCI are shown in Table 2. CI-AKI was most strongly associated with CKD. Patients who underwent PCI in urgent or emergent settings or presented with heart failure symptoms of New York Heart Association functional class of III or higher or ST-segment elevation myocardial infarction were also strongly correlated with CI-AKI. CI-AKI was likely to develop in patients with major bleeding with an odds ratio of 2.23 (95% confidence interval: 1.37 to 3.63; p = 0.001).
Five percent of patients who experienced CI-AKI died in the hospital, compared with 1% of patients without CI-AKI (p < 0.001) (Table 3). The rate of a composite of heart failure, cardiogenic shock, and death was also significantly higher in the CI-AKI group compared with the non–CI-AKI group (17% vs. 3%, p < 0.001). Patients who experienced CI-AKI had a much higher incidence of complications such as periprocedural myocardial infarction, heart failure, cardiogenic shock, new requirement for hemodialysis, need for blood transfusion, and major bleeding after PCI (31% vs. 11%, p < 0.001). Patients in whom CI-AKI developed after PCI had signiﬁcantly higher rates of major bleeding and blood transfusion compared with those in whom CI-AKI did not develop. Seven percent of patients who experienced CI-AKI required hemodialysis treatment compared with no patients in the non–CI-AKI group (p < 0.001).
Severity of Hb decrease and incidence of CI-AKI, major bleeding, and other complications
A significant positive association between the severity of the decrease in Hb and event rate was observed in our study population. During the 48 h after the procedure, the following events were observed: 315 CI-AKI cases overall (12%), 18 (6%) in group A, 57 (8%) in group B, 93 (11%) in group C, 83 (17%) in group D, and 64 (26%) in group E (p < 0.001) (Table 4, Fig. 2). Moreover, significantly increased rates of any complications and death were observed in those patients with more severe decrease in Hb level. Furthermore, the incidence of major bleeding significantly increased with a greater decrease in Hb drop (p < 0.001). On the other hand, patients with minor bleeding showed a nonsignificant trend of the development of CI-AKI (odds ratio = 2.22, 95% confidence interval: 0.62 to 7.93; p = 0.388) (Online Table 1).
The major findings of the present study were as follows: 1) patients who experienced periprocedural bleeding had a higher likelihood of the development of CI-AKI and 2) CI-AKI incidence correlated closely with bleeding severity. Patients in whom CI-AKI developed had poorer in-hospital outcomes, in both all-cause death and the composite of heart failure, cardiogenic shock, and death, than patients in whom CI-AKI did not develop.
An important finding of our study was that CI-AKI incidence and in-hospital complications were significantly associated with bleeding severity. Several studies have analyzed the relationship between CI-AKI incidence and baseline Cr level (1), CI-AKI incidence in elective versus emergent PCI settings (13); however, few studies have examined the relationship between bleeding severity and CI-AKI incidence. Previous studies have suggested an association between bleeding after PCI and late mortality. Applegate et al. (14) observed Thrombolysis In Myocardial Infarction bleeding grade to be an independent predictor of death at 1 year in a large population of 3,931 patients. Romaguera et al. (15) reported that patients with vascular complications and greater hematocrit decreases were at increased risk of death at 1 year.
Our study, however, conﬁrms the importance of the volume of periprocedural blood loss as a risk factor for patients experiencing CI-AKI. Even small absolute changes in Hb level were found to be associated with significantly increased rates of both CI-AKI and in-hospital complications. Does this change in Hb level always indicate bleeding? Previsdomini et al. (16) described the alteration of Hb level of nonbleeding patients with ACS during the first 24 h in the intensive care unit. In this series, the extent of Hb decrease was only 1.27 ± 1.00 g/dl, which is far from our bleeding threshold of 3 g/dl. Furthermore, Svensén et al. (17) described that hemodilution with fluid resuscitation requires a large amount of Ringer's solution, more specifically, a mean volume of 5.9 ± 0.8 l.
In the literature, an acute decrease in Hb is reported in severe AKI cases for which hemodialysis is required (18). In their report, a dramatic drop in Hb concentration occurred several days after the onset of AKI and was suggested due to decreased erythropoietin levels associated with AKI. However, in our studies, most of the patients in whom CI-AKI developed had a modest increase in Cr level, and only 7% of the patients in whom CI-AKI developed required hemodialysis. Moreover, we collected data on post-PCI Hb level within 72 h after the procedure (i.e., during the acute phase). The definition of CI-AKI used (i.e., an increase in serum Cr at 48 h after contrast media administration) renders it less likely that renal anemia contributed to a decrease in Hb.
Major bleeding may lead to hemodynamic compromise. It is important to distinguish whether CI-AKI is caused by hemodynamic compromise or by significant blood loss. In this study, we have excluded patients presenting with cardiogenic shock. Moreover, we performed 1-way ANOVA of systolic blood pressure (SBP), mean arterial pressure (MAP), and diastolic blood pressure (DBP) before and after PCI among groups A through E. Blood pressure change was also analyzed. There were no significant differences in SBP, MAP, or DBP before PCI among the 5 groups (p = 0.151, p = 0.158, and p = 0.213, respectively) (Online Table 2). Moreover, a change in SBP, MAP, and DBP also showed no significant differences among the 5 groups (p = 0.126, p = 0.736, and p = 0.592, respectively). Nonetheless, significantly increased rates of CI-AKI and bleeding events (p < 0.01 and p < 0.01, respectively) were observed in those patients with more severe decrease in Hb level. On the other hand, there were 49 cases (1.9%) in the study population of patients with a significant compromise of hemodynamics during PCI, which we have defined as cardiogenic shock. In patients with cardiogenic shock CI-AKI developed significantly more frequently compared with patients in whom it did not develop (7% vs. 1%, p < 0.001) (Table 3). Moreover, significantly increased rates of cardiogenic shock were observed in those patients with more severe decrease in Hb level (Table 4).
Another major risk factor of CI-AKI is the volume of contrast media. We performed 1-way ANOVA comparisons of contrast volume among the 5 groups A through E. There were no significant differences in contrast volume among the 5 groups (p = 0.519) (Online Table 2). Moreover, contrast volume adjusted for body mass index also showed no significant differences among the 5 groups (p = 0.079).
Against the background of the extensive literature on CI-AKI (1,2,5,6,19,20), which has been the subject of numerous comprehensive reviews (21,22), the present study provides additional clinical insight. In most catheterization laboratories, baseline Cr is currently assessed before PCI; therefore, interventional cardiologists can predict the risk of CI-AKI developing before the procedure, especially in elective cases. Those patients predicted to be at high risk of CI-AKI should undergo additional preventive measures and staged procedures, and perioperative blood loss should be minimized. In such cases, performing PCI via a transradial approach would be preferred because the risk of bleeding is significantly lower compared with the transfemoral approach (23,24). In a large Canadian registry, Vuurmans et al. (25) reported significantly higher incidence of CKD within 6 months of the cardiac procedure performed via the transfemoral approach compared with the transradial approach. Our study population also shows a higher incidence of CI-AKI with transfemoral intervention; however, most of the urgent or emergent cases were performed with transfemoral intervention; therefore, further confirmation is needed of whether performing PCI via the transradial approach decreases the incidence of CI-AKI.
Our study has the inherent limitations of its retrospective design, although the data were collected prospectively. Because the operators were not blinded to the clinical and laboratory information including baseline Cr and Hb levels, additional measures to prevent CI-AKI might have been undertaken in select patients. Due to the limited availability of data ﬁelds, data on the volume of contrast media, accurate periprocedural hydration volume, presence of proteinuria, urine output, and use of nephrotoxic medications were not available in our database. Although serum Cr levels were checked routinely after PCI, some patients in whom increased serum Cr developed after hospital discharge may have been missed. The type of AKI (hemodynamic, ischemic, nephrotoxic, or atheroembolic) could not be determined precisely. Because of the nature of both the patients and procedures performed, AKI was assumed to be multifactorial. Furthermore, because of methodological limitations inherent to retrospective registry analyses, our data cannot establish a definite etiological link between worsening renal function after PCI and the increased risk of major adverse cardiorenal events and death. Moreover, we could not demonstrate that bleeding precedes AKI in a vast majority of cases. Our findings should be confirmed in larger multicenter trials.
Periprocedural bleeding is significantly associated with CI-AKI, CI-AKI incidence correlates with bleeding severity, and patients who experience CI-AKI after PCI are at very high risk of in-hospital death.
The authors appreciate the contributions of all the investigators, clinical coordinators, and institutions involved in the JCD-KICS study.
The present study was supported by the Grants-in-Aid for Scientific Research KAKENHI (No. 25460630) and Pfizer Health Research Foundation. All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- acute coronary syndromes
- analysis of variance
- contrast-induced acute kidney injury
- chronic kidney disease
- diastolic blood pressure
- estimated glomerular ﬁltration rate
- mean arterial pressure
- percutaneous coronary intervention
- systolic blood pressure
- Received October 29, 2012.
- Revision received February 25, 2013.
- Accepted March 13, 2013.
- 2013 American College of Cardiology Foundation
- Rihal C.S.,
- Textor S.C.,
- Grill D.E.,
- et al.
- Best P.J.,
- Lennon R.,
- Ting H.H.,
- et al.
- Mehran R.,
- Aymong E.D.,
- Nikolsky E.,
- et al.
- Kshirsagar A.V.,
- Poole C.,
- Mottl A.,
- et al.
- Eikelboom J.W.,
- Mehta S.R.,
- Anand S.S.,
- Xie C.,
- Fox K.A.,
- Yusuf S.
- Ndrepepa G.,
- Berger P.B.,
- Mehilli J.,
- et al.
- Mehran R.,
- Rao S.V.,
- Bhatt D.L.,
- et al.
- McCullough P.A.
- Solomon R.,
- Dauerman H.L.
- Yang Y.J.,
- Kandzari D.E.,
- Gao Z.,
- et al.
- Vuurmans T.,
- Byrne J.,
- Fretz E.,
- et al.