Author + information
- Received September 12, 2011
- Accepted September 15, 2011
- Published online January 1, 2012.
- Hitinder S. Gurm, MD⁎,⁎ (, )
- Dean E. Smith, PhD, MPH⁎,
- Otavio Berwanger, MD†,
- David Share, MD‡,
- Theodore Schreiber, MD§,
- Mauro Moscucci, MD, MBA∥,
- Brahmajee K. Nallamothu, MD, MPH⁎,
- BMC2 (Blue Cross Blue Shield of Michigan Cardiovascular Consortium)
- ↵⁎Reprint requests and correspondence:
Dr. Hitinder S. Gurm, Department of Internal Medicine, Division of Cardiovascular Medicine, University of Michigan, 2A394, 1500 East Medical Center Drive, Ann Arbor, Michigan 48109-5853
Objectives The aim of this study was to examine the use of and outcomes associated with use of N-acetylcysteine (NAC) in real-world practice.
Background The role of NAC in the prevention of contrast-induced nephropathy (CIN) is controversial, leading to widely varying recommendations for its use.
Methods Use of NAC was assessed in consecutive patients undergoing nonemergent percutaneous coronary intervention from 2006 to 2009 in the Blue Cross Blue Shield of Michigan Cardiovascular Consortium, a large multicenter quality improvement collaborative. We examined the overall prevalence of NAC use in these patients and then used propensity matching to link its use with clinical outcomes, including CIN, nephropathy-requiring dialysis, and death.
Results Of the 90,578 percutaneous coronary interventions performed during the study period, NAC was used in 10,574 (11.6%) procedures, with its use steadily increasing over the study period. Patients treated with NAC were slightly older and more likely to have baseline renal insufficiency and other comorbidities. In propensity-matched, risk-adjusted models, we found no differences in outcomes between patients treated with NAC and those not receiving NAC for CIN (5.5% vs. 5.5%, p = 0.99), nephropathy-requiring dialysis (0.6% vs. 0.6%, p = 0.69), or death (0.6% vs. 0.8%, p = 0.15). These findings were consistent across many prespecified subgroups.
Conclusions Use of NAC is common and has steadily increased over the study period but does not seem to be associated with improved clinical outcomes in real-world practice.
Contrast-induced nephropathy (CIN) is a common complication of percutaneous coronary intervention (PCI) and responsible for excessive morbidity, mortality, and healthcare cost (1–5). Multiple strategies have been explored to reduce the occurrence of CIN after PCI (6), although the only proven strategies remain adequate peri-procedural hydration, avoidance of hyperosmolar contrast media, and minimization of contrast volume (7,8).
On the basis of the assumption that free radicals play a role in mediating CIN, a number of antioxidant agents have been studied in recent years with mixed results (9). The most notable of these is N-acetylcysteine (NAC), whose use was first described a decade ago (10). Yet multiple randomized clinical trials have demonstrated conflicting results about its clinical benefits, which has led to varying recommendations for its use (11). The recently presented and large ACT (Acetylcysteine for the prevention of Contrast-induced nephropathy Trial) demonstrated no apparent benefit for NAC with respect to occurrence of CIN (12). Although the ACT trial provided important data in this debate, even this large study is limited in its ability to examine high-risk patient subgroups, such as those receiving large doses of contrast, or “hard” clinical outcomes like nephropathy-requiring dialysis or death.
As such, there remains a paucity of data on the overall uptake of NAC in real-world practice and its association with outcomes. The aims of our study were 2-fold. First, we sought to identify the prevalence of NAC use in a broad and unselected population of patients undergoing PCI in the BMC2 (Blue Cross Blue Shield of Michigan Cardiovascular Consortium) a large multicenter quality improvement collaborative. Second, we assessed the impact of NAC use on reducing CIN, dialysis, and death, focusing on patient subgroups that might be at particularly high risk for these complications but in whom data are more sparse.
Data sources and study population
We included data from 31 hospitals in the BMC2. The details of the BMC2 and its data collection and auditing process have been described previously (13–15). Briefly, procedural data on all patients undergoing PCI at participating hospitals are collected with standardized data collection forms. Baseline data include clinical, demographic, procedural, and angiographic characteristics as well as medications used before, during, and after the procedure and in-hospital outcomes. All data elements have been prospectively defined, and the protocol is approved by local institutional review boards at each hospital. In addition to a random audit of 2% of all cases, medical records of all patients undergoing multiple procedures or coronary artery bypass grafting and of patients who died in the hospital are reviewed routinely to ensure data accuracy.
The study population for this analysis included all consecutive patients who underwent nonemergent PCI between January 1, 2006, and December 31, 2009. We excluded patients undergoing PCI for ST-segment elevation myocardial infarction, cardiogenic shock, or cardiac arrest, because the urgency inherent in performing PCI for these indications precludes administration of oral NAC. We also excluded patients who were already receiving dialysis at the time of their procedure or those who died in the catheterization laboratory because they did not survive long enough to possibly benefit from NAC. N-acetylcysteine use was defined as administration of at least 1 oral dose of NAC within the 24 h before the procedure. The dose and duration of NAC administration was as per the local institutional protocol and operator preference.
Our study had 3 endpoints for this analysis: CIN, nephropathy-requiring dialysis, and in-hospital death. The CIN was defined as impairment in renal function resulting in ≥0.5 mg/dl absolute increase in serum creatinine from baseline. Baseline creatinine was collected within 1 month of the procedure. Among patients who had multiple assessments of serum creatinine in the 30 days before the procedure, the value closest to the time of the procedure was considered as the baseline value. Peak creatinine was defined as the highest value of creatinine in the week after the procedure and was ascertained as per local clinical practice. A follow-up creatinine was collected at least 1 day after the procedure but varied, depending on length of stay. Nephropathy-requiring dialysis was defined as new, unplanned need for hemodialysis during the hospital stay for patients, due to worsening of renal function after PCI. In-hospital death was obtained from the medical records.
We separated patients into 2 categories: those treated with NAC pre-procedurally, and those who were not. Continuous variables were expressed as mean ± SD, and discrete variables were expressed as frequency counts and percentages. Continuous variables were analyzed with the Student t test, with the Wilcoxon rank-sum test used for skewed variables. Differences in discrete variables between patients treated with NAC and those who did not receive it were evaluated by the chi-square-test and Fisher exact test. Trends in NAC use over time in all patients and in different patient subgroups were assessed with the Cochran Armitage test.
We used propensity matching to account for the nonrandom use of NAC (16,17). A propensity score for receiving NAC was calculated with a nonparsimonious logistic regression model created from several pre-procedural patient characteristics. These patient characteristics included age, sex, baseline glomerular filtration rate (GFR), smoking status, history of hypertension, diabetes, prior myocardial infarction, congestive heart failure, prior coronary revascularization (PCI or coronary artery bypass grafting), atrial fibrillation, peripheral vascular disease, history of stroke or transient ischemic attack, and chronic obstructive pulmonary disease. Greedy matching technique was then used to match a patient receiving NAC to a patient with the nearest propensity score who did not receive NAC to permit adequate comparison of similar patients (18). Matching adequacy was confirmed by calculating the standardized differences for baseline variables among the matched subset, with a standardized difference of <10% suggesting adequate matching. The study endpoints were then compared within this propensity-matched cohort. Random effect hierarchical models were fitted to control for clustering and variation by hospital. Separate models were developed for CIN, nephropathy-requiring dialysis, and in-hospital death.
We further evaluated differences in outcomes among several key, high-risk patient subgroups with each subgroup defined by the presence or absence of a single high-risk characteristic. These subgroups included patients with abnormal renal function at baseline, diabetes, and anemia and those treated with high contrast volume and by the specific type of contrast media employed (iso-osmolar contrast or low osmolar contrast media). Glomerular filtration rate was estimated with the Modification of Diet in Renal Disease equation (19). Adjustment for multiple comparisons was performed during these additional analyses with the false discovery rate method (20). The SAS software (version 9.1, SAS, Cary, North Carolina) was used for all analysis.
Our study population comprised 90,578 PCI procedures with NAC used in 10,574 (11.64%) patients. A slight increase in the proportion of patients treated with NAC was noted over time, increasing from 10.1% to 13.0% over the study period. Use of NAC was common in patients with abnormal renal function at baseline, including 58.6% of patients with a serum creatinine >1.5 mg/dl and 68.7% of those with a serum creatinine >2 mg/dl. Baseline characteristics of patients treated with NAC compared with those not receiving NAC are described in Table 1. Patients treated with NAC were older, more likely to have worse renal function at baseline, and had more comorbidities, including hypertension, diabetes, anemia, lower ejection fraction, prior coronary artery disease, and prior coronary revascularization. The average dose of contrast media used was slightly lower in patients treated with NAC, although procedure duration as assessed by total fluoroscopic time was longer. Significant variation in NAC use was noted across the different hospitals, ranging from 1.4% to 33.0% of all treated patients.
By comparison, the propensity-matched cohorts were well-balanced, with no major difference in the baseline characteristics of patients treated with NAC compared with those not receiving NAC (Table 2). The model developed to predict the propensity score had a C statistic of 0.86, suggesting high discrimination. Overall, unadjusted and adjusted outcomes for the study cohort are described in Table 3. In unadjusted analyses, patients treated with NAC had a significantly higher incidence of CIN, nephropathy-requiring dialysis, and in-hospital death. However, no differences in the incidence of CIN, nephropathy-requiring dialysis, and in-hospital death were observed between patients treated with NAC compared with those who did not receive it in the propensity-matched analysis (Fig. 1, Table 3). When we examined the occurrence of CIN and nephropathy-requiring dialysis in different patient subgroups, a lower incidence of CIN was noted in patients with GFR <30 ml/min/1.73 m2, but this was not significant after adjusting for multiple comparisons (Fig. 2A). No difference in outcomes was observed in any other subgroup on the basis of baseline renal function, presence or absence of diabetes or anemia, the contrast dose, or the specific type of contrast media employed (iso-osmolar contrast or low osmolar contrast media) (Fig. 2). Similarly, when patients were categorized as those with a baseline serum creatinine <1.5 mg/dl or ≥1.5 mg/dl, no difference in the incidence of CIN (4.32% vs. 3.98%, p = 0.41, among those with serum creatinine <1.5 mg/dl; and 7.14% vs. 7.76%, p = 0.35, among those with serum creatinine ≥1.5 mg/dl) or nephropathy-requiring dialysis (0.17% vs. 0.13%, p = 0.57, among those with serum creatinine <1.5 mg/dl; and 1.29% vs. 1.29%, p = 0.99, among those with serum creatinine ≥1.5 mg/dl) was observed among patients treated with NAC compared with the matched set of patients who were not treated with NAC.
In a sensitivity analysis, no difference in the incidence of CIN or nephropathy-requiring dialysis was noted among the propensity-matched cohort in patients with a post-procedural length of stay of 2 days (n = 2,180, CIN 3.23% with NAC vs. 4.42%, p = 0.15, and nephropathy-requiring dialysis 0.17% vs. 0.30%, p = 0.53) or among those with a length of stay of 3 days or longer (n = 3,383, CIN 20.54% with NAC vs. 20.70%, p = 0.91, and nephropathy-requiring dialysis 2.95% vs. 2.66%, p = 0.60).
The key finding of our study is that, in our study population, NAC is commonly used in patients undergoing PCI, although it seems to have no consistently beneficial effects on in-hospital outcomes. The results of our study add to and extend the findings of the recently presented ACT trial in 2 important ways. First, the large size of our study cohort allowed us to examine the impact of NAC on hard outcomes, such as nephropathy-requiring dialysis and death. Second, we were able to assess outcomes among subgroups, such as those with advanced renal dysfunction or those treated with large volumes of contrast media that were under-represented in the ACT trial.
Importantly, in these additional subgroup analyses, we identified a possible reduction in CIN in patients with a GFR <30 ml/min/1.73 m2, although there was no reduction in nephropathy-requiring dialysis. A possible explanation is that this simply reflects a chance finding, because there was no trend for significance after adjusting for multiple comparisons. Given the consistently null results observed in our study, it is unlikely that NAC has any clinically relevant effect in patients undergoing PCI, but this isolated group might be worthy of further study. Until additional studies are performed, however, we would not recommend its use on the basis of this isolated result.
The widespread use of NAC especially in patients with abnormal renal function despite the lack of definitive evidence is noteworthy. The evidence base supporting NAC is inconsistent, yet some authorities advocate its use on the assumption of low cost, possible benefit, and lack of harm (21). Although a few trials and meta-analysis have demonstrated dramatic benefits, studies with stronger methodology have been mostly negative (22). Furthermore, the field has been dominated by multiple inadequately powered trials, and it took 10 years for the first adequately powered trial to be completed. The results of our work are strikingly similar to that of the recently presented ACT trial, which found little difference in outcome of patients randomized to NAC compared with placebo.
Although observational data in general cannot supplant clinical trials, large studies such as ours can be helpful in complementing randomized trials, particularly when there are understudied groups of patients where data might be less adequate. The results of our study combined with those of the ACT trial suggest that there is no role for NAC for prophylaxis of CIN in the vast majority of patients and across a number of high-risk subgroups. Given the prior widespread use of this agent, a move away from this practice should translate into a reduction in healthcare resources and allow practitioners to focus instead on proven prophylactic strategies. Although often dismissed as a low-cost therapy, the administration of NAC in approximately 10% of patients can add significantly to healthcare cost. Even though the drug itself is inexpensive, the time, effort, and opportunity costs required for setting up systems to identify high-risk patients and then to dispense it are not inconsequential.
The BMC2-PCI registry is a regional database with an active focus on multi-centric quality improvement and might or might not be representative of the wider population of patients undergoing PCI in the United States. Nevertheless, the registry is comprehensive in that it includes almost all PCIs performed in the state of Michigan. The registry does not track data on the total dose of NAC administered, and no dose response analysis could be performed. The study was also subject to unmeasured residual confounding, despite propensity-matched analyses that might influence any of the noted associations. Serum creatinine was not checked in a systematic fashion, and not all patients had serum creatinine checked within the 48- to 72-h window that corresponds with peak serum creatinine elevation in patients with CIN. However, in a sensitivity analysis restricted to patients with longer length of stay, no benefit of NAC could be discerned. Finally, the term “CIN” implies that contrast media are solely responsible for all cases of renal dysfunction occurring in this population, although this is probably incorrect. We elected to use this term because there is no widely accepted alternative.
N-acetylcysteine is commonly used for prevention of CIN in patients undergoing PCI. No clinical benefit was noted in association with NAC use. The results of our study combined with those of the ACT trial suggest that this practice should be abandoned.
The authors are indebted to all the study coordinators, investigators, and patients who participated in the Blue Cross Blue Shield of Michigan Cardiovascular Consortium.
The Blue Cross Blue Shield of Michigan Cardiovascular Consortium registry is funded by Blue Cross Blue Shield of Michigan. The sponsor had no role in analysis, study design, or decision to publish these results. Dr. Gurm receives research funding from Blue Cross Blue Shield of Michigan and the National Institutes of Health. Dr. Share is employed part-time by Blue Cross Blue Shield of Michigan. Mauro Moscucci has received salary support from the Blue Cross Blue Shield of Michigan.
All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- contrast-induced nephropathy
- glomerular filtration rate
- percutaneous coronary intervention
- Received September 12, 2011.
- Accepted September 15, 2011.
- American College of Cardiology Foundation
- Rihal C.S.,
- Textor S.C.,
- Grill D.E.,
- et al.
- ACT Trial Investigators
- Moscucci M.,
- Rogers E.K.,
- Montoye C.,
- et al.
- Gurm H.S.,
- Smith D.E.,
- Collins J.S.,
- et al.
- Parsons L.
- Benjamini Y.,
- Hochberg Y.
- Poldermans D.,
- Bax J.J.,
- Boersma E.,
- et al.,
- Task Force for Preoperative Cardiac Risk Assessment and Perioperative Cardiac Management in non-cardiac surgery of the European Society of Cardiology (ESC) and European Society of Anaesthesiology (ESA)