N-acetylcysteine and contrast-induced nephropathy: a meta-analysis of 13 randomized trials.

BACKGROUND: Contrast-induced nephropathy (CIN) following coronary angiography increases morbidity and mortality. Randomized trials of small sample size have evaluated whether N-acetylcysteine (NAC) prevents CIN in patients with renal dysfunction.

METHODS: To conduct a meta-analysis of the randomized trials the following databases were searched: MEDLINE (1966-2003), Cochrane Controlled Trials Register, ACP Journal Club online, published abstracts presented at the major cardiology and nephrology meetings, references from reviews. Two authors independently evaluated all relevant randomized trials. Eligibility criteria were (1) randomized placebo controlled trials of NAC, (2) patients with impaired renal function (creatinine >1.2 mg/dL) undergoing coronary angiography, (3) patients receiving intravenous fluids and low-osmolarity nonionic contrast media, (4) the primary outcome was CIN (increases in creatinine of either at least 0.5 mg/dL or 25% from baseline to 48 hours). Of 589 trials reviewed 3 disagreements were easily resolved by mutual discussion and 13 were selected. Data extraction included patient characteristics, intravenous fluid regimen, type and dose of contrast media, dosing regimen, creatinine at baseline and 48 hours and CIN requiring dialysis.

RESULTS: Four of the 13 trials reported statistically significant results. In meta-analysis of the 13 trials, which included 1892 patients, the RR was 0.68 (95%CI, 0.46-1.01). The addition of the trial of patients undergoing computerized tomography, which had formulated the hypothesis, yielded a statistically significant reduction (RR 0.64 [95%CI 0.42-0.96]) as did an earlier meta-analysis of 7 trials.

CONCLUSIONS: Our meta-analysis of the most currently available randomized data concerning NAC before coronary angiography to prevent CIN in patients with impaired renal function is neither conclusive nor provides proof beyond a reasonable doubt to influence clinical practice and public policy. The intervention has minimal toxicity but the width of the 95% CI remains compatible with a range from a large benefit to none at all. In addition, the trials used change in creatinine as the measure of outcome. Further randomized trials of large sample size and with clinical outcomes will add importantly relevant information to the totality of evidence and allow the most rational clinical decisions for individual patients as well as policy decisions for the health of the general public.