RSS Feeds
Is There Enough Evidence to Support Use of N-Acetylcysteine in Contrast-Induced Nephropathy?
- Aine M. Kelly, MD, MS;
- Ben Dwamena, MD; and
- Ruth C. Carlos, MD, MS
IN RESPONSE:
We thank Dr. Trivedi for detecting 2 errors in our article. First, we included 26 studies of N-acetylcysteine, not 30 as indicated in Figure 1. Second, we erroneously cited a review related to the selection of contrast media (1)—we intended to cite a 1994 study by Solomon and colleagues (2). In addition, we agree with Dr. Trivedi that substantial heterogeneity of studies evaluating N-acetylcysteine makes the conclusions of our analysis far from definitive. As stated in our article, head-to-head trials of available agents are needed to define the most effective and safe strategy of preventing contrast-induced nephropathy. Finally, the editors have published a correction about the Editors' Notes accompanying our article (3).
Dr. Gonzalez and colleagues note that previous meta-analyses have examined the efficacy of N-acetylcysteine in preventing contrast-induced nephropathy and that some of these, including their own in 2007 (4), have reached different conclusions than ours. Their 2007 analysis included 22 trials and suggested that most of the apparent benefit of N-acetylcysteine was attributable to 4 small, relatively early trials (5–8). Our meta-analysis was larger, including 26 studies, and was less prone to the influence of these 4 trials. We acknowledge that creatinine level is an imperfect surrogate outcome measure for nephropathy, that protection against nephropathy based on change in creatinine level has not been confirmed by using other measures, and that the mechanism by which N-acetylcysteine might prevent contrast-induced nephropathy remains uncertain. Although Dr. Gonzalez and colleagues are correct in noting that anaphylactoid reactions occurred in studies of intravenous N-acetylcysteine treatment of acetaminophen poisoning, the doses of N-acetylcysteine used for acetaminophen overdose are larger and the duration of therapy is longer (150 mg/kg of body weight initial bolus dose and 50 mg/kg over 4 hours, and 100 mg/kg for 16 hours). At the N-acetylcysteine doses and durations used in the trials we cited, no major adverse effects were reported. However, it is wise to advise caution in the use of N-acetylcysteine in hospitalized patients. As noted in the discussion of our results, we agree that the efficacy of N-acetylcysteine for preventing contrast-induced nephropathy remains unproven and that creatinine measurement is a surrogate value.
Dr. Ferrante and colleagues raise several issues about our methods. Although we explored the pooling of treatment effects from the 41 included trials to derive a single summary estimate for the effectiveness of the use of a renoprotective agent in reducing contrast-induced nephropathy, we refrained from reporting these analyses after reviewers noted substantial problems with the pooling of these very heterogeneous studies and the uncertain interpretation of the summary estimate in clinical terms. These exploratory analyses suggested that patient age and baseline creatinine level influenced the effects of renoprotective agents on contrast-induced nephropathy and that diabetes mellitus, patient sex, and hypertension did not. However, use of summary measures, such as average age or proportion of female study participants, as independent variables in meta-regression models is subject to potential ecological biases. Pooled analyses using individual patient data and additional high-quality trials are needed to reliably explore these sources of heterogeneity. We did not look into whether the dose, timing, and duration of N-acetylcysteine contributed to heterogeneity, and we agree that this would be interesting to pursue in future analyses. We used the method of DerSimonian and Laird to assess overall and subgroup summary risk ratios and I2 to assess overall and subgroup summary risk ratios and to derive the P value reported in Figure 2 of our article.
Dr. Berwanger disagrees with our conclusion that “the use of … N-acetylcysteine is reasonable in high-risk patients who are to receive large or repeated volumes of contrast agents.” As noted in our article, we agree that clinicians need to understand that evidence supporting N-acetylcysteine's effectiveness comes from studies that have measured surrogate end points, have been of generally poor quality, and have found heterogeneous results. Yet, clinicians must make decisions with the evidence presently on hand. Although far from definitive, this evidence is more encouraging for the effectiveness of N-acetylcysteine than that for other agents used to prevent contrast-induced nephropathy. High-quality randomized, controlled trials evaluating clinical outcomes may alter our current conclusions about the effectiveness of N-acetylcysteine when they become available.
Aine M. Kelly, MD, MS
Ben Dwamena, MD
Ruth C. Carlos, MD, MS
University of Michigan
Ann Arbor, MI 48109
Article and Author Information
-
Potential Financial Conflicts of Interest: None disclosed.
