Beyond the AJR: Why Was N-Acetylcysteine Considered an Effective Pretreatment for Contrast-Induced Acute Kidney Injury for So Long?

For decades, the antioxidant N-acetylcysteine (NAC) was considered an effective pretreatment to reduce the risk of contrast-induced acute kidney injury (CI-AKI) after exposure to iodinated contrast material. However additional research eventually refuted this claim, and NAC is no longer recommended to prevent CI-AKI. To better understand the origins of these discrepant results, Magner et al. [1] performed a meta-analysis of 101 randomized controlled trials of NAC prophylaxis. They found that NAC pretreatment did reduce the risk of CI-AKI when all studies were examined (random-effects model pooled OR, 0.72; 95% CI, 0.63–0.82); however, the significant study heterogeneity observed (I²=39; p< .001) called into question the validity of these results. When restricted to larger studies (> 500 participants [three studies] or > 100 events [six studies]), study heterogeneity was eliminated (I²=0), and no residual benefit of NAC prophylaxis was observed (> 500 participants: pooled OR, 1.03 [95% CI, 0.89–1.18]; > 100 events: pooled OR, 1.03 [95% CI, 0.90–1.16]). Results of funnel plot analysis suggested the absence of smaller NAC trials that did not show an effect. When the authors performed an imputed analysis of these missing studies, the result shifted toward null. The authors concluded that numerous smaller, more heterogeneous NAC studies erroneously drove initial conclusions that NAC was an effective CI-AKI pretreatment; larger studies with minimal heterogeneity instead showed that NAC was not an effective prophylaxis against CI-AKI. The authors also found that study heterogeneity and potential bias varied with outcomes. Compared with biomarker analysis studies with significant heterogeneity, studies examining more clinically relevant outcomes did not exhibit this heterogeneity (I²=0); NAC did not reduce risk of dialysis (pooled OR, 0.86; 95% CI, 0.69–1.12) or all-cause death (pooled OR, 0.93; 95% CI, 0.77–1.11). The authors thus argue for a transition from a biochemical definition of CI-AKI (e.g., changes in serum creatinine or other renal bio-markers) to a definition based on relevant clinical outcomes (e.g., postcontrast incidence of dialysis or mortality).