Effects of Intra-Arterial and Intravenous Iso-Osmolar Contrast Medium (Iodixanol) on the Risk of Contrast-Induced Acute Kidney Injury: A Meta-Analysis

Abstract

BACKGROUND: The iso-osmolar contrast agent iodixanol may be associated with a lower incidence of contrast-induced acute kidney injury (CI-AKI) than low-osmolar contrast media (LOCM), but previous meta-analyses have yielded mixed results. Objectives: To compare the incidence of CI-AKI between iodixanol and LOCM. METHODS: Studies were identified from literature searches to December 2009, clinicaltrials.gov, and conference abstracts from the past 2 years including 2010. Only prospective, randomized comparisons between iodixanol and LOCM with CI-AKI [increase in serum creatinine (sCr) ≥0.5 mg/dl or ≥25% from baseline, as defined in the trial] as a primary and/or secondary endpoint and a Jadad score ≥2 were included. A random-effects model was used to obtain pooled relative risks (RRs) for CI-AKI in analyses based on route of administration [intra-arterial (IA) or intravenous (IV)], definition of CI-AKI, and timing of sCr measurements. RESULTS: 145 potential articles were identified, of which 25 were included in the meta-analysis. Following IA administration (n = 19), the RR for CI-AKI (≥0.5 mg/dl definition) with iodixanol, compared with LOCM, was 0.462 [95% confidence interval (CI): 0.272-0.786, p = 0.004, 15 studies]. Using the ≥25% definition, there was a lower incidence of CI-AKI with iodixanol versus LOCM, but the difference was not statistically significant (RR: 0.577, 95% CI: 0.297-1.12, p = 0.104, 11 studies). In the IV trials, there was no significant difference in the incidence of CI-AKI using either definition (≥0.5 mg/dl definition: RR: 0.967, 95% CI: 0.188-4.972, p = 0.968, 3 trials; ≥25% definition: RR: 0.656, 95% CI: 0.316-1.360, p = 0.257, 4 trials). CONCLUSIONS: IA but not IV administration of iodixanol is associated with a significantly lower risk of CI-AKI than LOCM.

Figures

Fig. 1

Study selection.

Fig. 2

Meta-analysis of the incidence of CI-AKI (defined as ≥0.5 mg/dl increase in sCr from baseline) in trials reporting this outcome comparing IA iodixanol (IOCM) with LOCM.

Fig. 3

Meta-analysis of the incidence of CI-AKI (defined as ≥25% increase in sCr from baseline) in trials reporting this outcome comparing IA iodixanol (IOCM) with LOCM.

Fig. 4

Mean changes in sCr in studies with standardized sCr determinations. The effect size was expressed as Hedges’ g[22].

Fig. 5

Compilation of pooled odds ratios for IA, IV, and mixed IA and IV meta-analyses of the incidence of CI-AKI (defined as ≥0.5 mg/dl increase in sCr vs. baseline) demonstrating a leftward shift in pooled estimates moving from IV, to mixed IV/IA, and IA trials. Pooled odds ratios from meta-analyses by Heinrich et al. [9], Reed et al. [10] and From et al. [11].

Fig. 6

Compilation of pooled odds ratios for IA, IV, and mixed IA and IV meta-analyses of the incidence of CI-AKI (defined as ≥25% increase in sCr vs. baseline) demonstrating a leftward shift in pooled estimates moving from IV, to mixed IV/IA, and IA trials. Pooled odds ratios from meta-analyses by Heinrich et al. [9], Reed et al. [10] and From et al. [11].