Acetylcysteine for prevention of contrast nephropathy: meta-analysis

ARTICLES

Acetylcysteine for prevention of contrast nephropathy: meta-analysis

Rainer Birck, Stefan Krzossok, Florian Markowetz, Peter Schnülle, Fokko J van der Woude, Claude Braun

Summary

Introduction

Background Contrast nephropathy is associated with increased in-hospital morbidity and mortality and leads to extension of hospital stay in patients with chronic renal insufficiency. Acetylcysteine seems to be a safe and inexpensive way to reduce contrast nephropathy. We aimed to assess the efficacy of acetylcysteine to prevent contrast nephropathy after administration of radiocontrast media in patients with chronic renal insufficiency. Methods We did a meta-analysis of randomised controlled trials comparing acetylcysteine and hydration with hydration alone for preventing contrast nephropathy in patients with chronic renal insufficiency. The trials were identified through a combined search of the BIOSIS+/RRM, MEDLINE, Web of Science, Current Contents Medizin, and The Cochrane Library Databases. We used incidence of contrast nephropathy 48 h after administration of radiocontrast media as an outcome measure. Findings Seven trials including 805 patients were eligible according to our inclusion criteria and were analysed. Overall incidence of contrast nephropathy varied between 8% and 28%. Since significant heterogeneity was indicated by the Q statistics (p=0·016) we used a random-effects model to combine the data. Compared with periprocedural hydration alone, administration of acetylcysteine and hydration significantly reduced the relative risk of contrast nephropathy by 56% (0·435 [95% CI 0·215–0·879], p=0·02) in patients with chronic renal insufficiency. Meta-regression revealed no significant relation between the relative risk of contrast nephropathy and the volume of radiocontrast media administered or the degree of chronic renal insufficiency before the procedure.

Radiocontrast media can lead to a reversible decline of excretory renal function that starts soon after administration, with an incidence of 10–30% in selected high-risk patients. Generally, this form of acute renal failure follows a benign course and only rarely necessitates use of dialysis.1–3 Nevertheless, use of radiocontrast media has been associated with increased in-hospital morbidity, mortality, and costs of medical care, and long admissions, especially in patients needing dialysis.4–7 Unfortunately, the best treatment to prevent contrast nephropathy remains to be established.1 Trials of diuretics, dopamine, calcium-channel blockers, atrial natriuretic peptides, aminophylline, and endothelin antagonists have yielded contrasting results.1 Only periprocedural hydration is widely accepted to prevent contrast nephropathy.1,8,9 Radiocontrast media reduces renal excretory function by altering renal haemodynamics and by exerting direct tubulotoxic effects.1 Although the exact pathophysiology of contrast nephropathy is not understood fully, results of work in animals have implicated reactive oxygen species in the pathogenesis of this disorder.10,11 Prevention of contrast nephropathy might be possible with the antioxidative agent acetylcysteine.12,13 In a randomised controlled trial14 in patients with chronic renal insufficiency, acetylcysteine significantly reduced incidence of contrast nephropathy after intravenous administration of radiocontrast media for elective CT examinations. However, results have contrasted in six other randomised controlled trials15–20 in which the effectiveness of acetylcysteine was assessed in patients with chronic renal insufficiency. We aimed to assess whether prophylactic use of acetylcysteine reduces incidence of contrast nephropathy in patients with renal insufficiency.

Methods Interpretation Compared with periprocedural hydration alone, acetylcysteine with hydration significantly reduces the risk of contrast nephropathy in patients with chronic renal insufficiency. The relative risk of contrast nephropathy was not related to the amount of radiocontrast media given or to the degree of chronic renal insufficiency before the procedure. Lancet 2003; 362: 598–603 See Commentary page 589

Fifth Department of Medicine, University Hospital Mannheim, Mannheim, Germany (R Birck MD, S Krzossok MD, P Schnülle MD, F J van der Woude MD, C Braun MD); and Max Planck Institute for Molecular Genetics, Berlin, Germany (F Markowetz) Correspondence to: Dr Rainer Birck, Fifth Department of Medicine, University Hospital Mannheim, Theodor-Kutzer-Ufer 1-3, D-68135 Mannheim, Germany (e-mail: [email protected])

598

We did this meta-analysis according to a predetermined protocol following the recommendations of the quality of reporting of meta-analysis (QUOROM) statement.21 To identify relevant studies, we searched BIOSIS+/RRM (1989 onwards), MEDLINE (1966 onwards), Web of Science (1997 onwards), the Current Contents Medizin (current contents of medical journals publishing in German language), and The Cochrane Library (1996 onwards). For the first four databases we used the keywords contrast nephropathy, N-acetylcysteine, acetylcysteine, renal failure, radiocontrast media, and randomised controlled trial. Various combinations of these terms were used depending on the requirements of the database. We searched The Cochrane Library using the terms contrast nephropathy, N-acetylcysteine, acetylcysteine, and radiocontrast media. We also searched PubMed with the related articles feature. The reference lists of identified papers and published reviews were checked and the proceedings from major cardiology and nephrology meetings within the past 5 years were screened. We included items that were available to us on Feb 5, 2003.

THE LANCET • Vol 362 • August 23, 2003 • www.thelancet.com

For personal use. Only reproduce with permission from The Lancet

ARTICLES

Patients Exclusion criteria

Relevant details

Radiocontrast procedure

Intervention

Primary endpoint*

Other remarks

Study Adamian34

57

Quasi-randomised (consecutive patients)

Coronary interventions NAC 600 mg orally NIRCM (iodixanol) twice daily + intravenous hydration

SCrea 25% increase Published as of baseline at 48 h abstract after procedure CN significantly reduced in NAC treated patients

Bader32

36

Randomised, but No hydration given no placebo group since pre-existing (NAC vs theophyllin) volume overload inclusion criterion

Angiography or CT scan NIRCM (iopromide, iohexol)

NAC 600 mg intravenous 30–45 min before procedure and 24 h later

Mean change in GFR (mL/min)

Mohayar35

55

Not randomised (case/control study)

Coro

NAC three times Mean change of daily before SCrea 48 h after procedure+ procedure intravenous hydration

Published as abstract. Significant benefit in NAC treated patients

Tepel and van der Giet33, 36

27

Post-hoc subgroup Only diabetics with CT scan analysis of CRI included NIRCM (iopromide) reference 14†

ACC 600 mg twice SCrea 0·5 at 48 h daily the day before after procedure and on the procedure day+ 0·45% saline 12 h before and after

Published as abstract. CN significantly reduced in NAC treated patients

Randomised controlled, but CRI not primary inclusion criteria

NAC 600 mg orally twice daily the day before and on the procedure day+ 0·9% saline (1 mL/kg/h 1–2 h before and 24 after RCM)

Vallero31

100

Post-hoc analysis Coro, PTCA of 20 patients with NIRCM (iodixanol) SCrea>12 mg/L

SCrea 0·5 or 33% increase of baseline at 48 h after procedure

Published as abstract Theophyllin significantly better than NAC

Original paper in Italian. No significant effect of NAC

NAC=acetylcysteine. CN=contrast nephropathy. SCrea=serum creatinine. NIRCM=non-ionic radio contrast media. Coro=coronary angiography. *mg/L unless stated otherwise. †Personal communication M Tepel.

Table 1: Characteristics of studies not eligible for meta-analysis

We included studies in all languages irrespective of masking when the following criteria were met: planned as a prospective, randomised, controlled trial; published as an article or abstract; included patients with chronic renal insufficiency receiving contrast media intravenously or intra-arterially for diagnostic or therapeutic procedures; use of acetylcysteine to prevent contrast nephropathy; and development of acute renal failure after administration of contrast media as a primary outcome. Trials that were retrospective, non-randomised or quasi-randomised, or compared different preventive measures without placebo control group were excluded. Two independent individuals (RB and CB) did the search and confirmed the eligibility of the identified trials. The same authors extracted the data independently of each other, and disagreements were resolved by discussion. The outcome measure was development of contrast nephropathy as defined by an increase in serum creatinine of at least 44·2 mol/L or of 25% from baseline values 48 h after administration of radiocontrast media. The methodological quality of the included trials was independently scored by RB and CB by the validated Jadad 5 point scale. The scale consists of three items describing randomisation (0–2 points), masking (0–2 points), and dropouts and withdrawals (0–1 points) in the report of a randomised-controlled trial.22 Higher scores indicate better reporting. Statistical analysis For every trial we calculated the relative risk for the primary outcome. The Q statistic was calculated to assess if significant heterogeneity was present between the included trials. This statistic tests the null hypothesis that the underlying effect measured by the pooled studies is equivalent. For p values less than 0·1 this assumption was deemed invalid.23 Since the Q statistic indicated that significant heterogeneity was present, we used the random-

effects model to combine the effect sizes of the included studies.24 The relation between the volume of radiocontrast media administered or degree of chronic renal insufficiency before the procedure and the relative risk of contrast nephropathy was investigated with metaregression. The regression model relates the treatment effect to study-level covariates with a within-study and an additive between-studies component of variance.25 To test the robustness of our findings to different assumptions, we also combined the effect sizes of all seven studies with a fixed-effects model.26 The effect of the putatively different pre-existing risk profiles between patients receiving radiocontrast media for cardiovascular procedures and patients in Tepel’s study,14 which included patients scheduled for elective CT examinations, was also assessed by excluding the latter from analysis. Furthermore, we recalculated the pooled relative risks after weighting the studies with a quality weight that was the product of the precision (ie, inverse variance) and the quality score (according to the Jadad scale).27 We assessed for presence of publication bias with a funnel plot for asymmetry28—a scatter plot of studies that relates the magnitude of the treatment effect against a measure of its precision. For formal statistical testing we used an adjusted rank correlation and a regression asymmetry test.29,30 A p value of less than 0·05 was judged significant with the exception of the Q statistics, in which a significance level of less than 0·1 was chosen. All statistical analyses were done with Stata version 6.0.

Results We identified 13 references describing studies in which acetylcysteine was used for prevention of contrast nephropathy in patients with chronic renal insufficiency, all of which were published between 2000 and 2003. Eight were full-length reports in peer-reviewed journals,14–20,31 one was non-English,31 and five were abstracts.32–36 Some of

THE LANCET • Vol 362 • August 23, 2003 • www.thelancet.com

For personal use. Only reproduce with permission from The Lancet.

599

ARTICLES

Patients Study (Jadad score) Tepel14 (1)

Allaqaband18 (3)

Briguori16 (1)

Patient details*

Procedure and agent

Acetylcysteine†

83

SCrea 24 (13) Age 66 years (45–66) DM 27 (33%)

Elective CT examination 75 mL iopromide intravenous

85

SCrea 21 (8) Age 71 years (10) DM 43 (50%)

Coro±PTCA Periph angiography±PTA 125 mL ioversol/dixanol

600 mg twice daily SCrea 5 the day before and on the procedure day+ 0·45% saline 12 h before and after 600 mg twice daily SCrea 5 the day before and on the procedure day+ 0·45% saline 12 h before and after

183

SCrea 15 (NA) Age 64 years (NA) DM 70 (38%)

Coro±PTCA Periph angiography 187 mL iopromide

Diaz-Sandoval17 (4)

54

SCrea 16 (3) Age 73 years (1) DM 11 (21%)

Elective cardiac catheterisation 184 mL ioxilan

Durham14 (3)

79

SCrea 23 (NA) Age 70–72 years DM 38 (48%)

Elective/acute coro 77–85 mL iohexol

Shyu15 (2)

121

SCrea 28 (8) Age 70 years (7) DM 77 (64%)

Coro±PTCA 118 mL iopamidol

Kay19 (5)

200

SCrea 14 (4) Age 68 years (7) DM 75 (38%)

Coro±PTCA 139 mL iopamidol

Primary endpoint‡ Other remarks No dialysis necessary Overall CN incidence 12%

Three group study: control, fenoldopam, NAC. Temporary dialysis needed in two patients from NAC group (1·6%)*** Overall CN incidence 16% 600 mg twice daily SCrea 25% or need Temporary dialysis the day before and on for dialysis first needed in one patient procedure day+ 5 days after from PLA group 0·45% saline 12 h procedure (0·5%). Post-hoc before and after subgroup analysis shows significant benefit in patients receiving less than 140 mL RCM Overall CN incidence 9% 600 mg twice daily SCrea 5 or 25% No dialysis diluted in ginger ale necessary*  4 doses (1 dose Overall CN before and 3 doses incidence 28% after procedure) + 0·45% saline for 2–12 h before and 12 h after 1200 mg 1 h before SCrea 5 Temporary dialysis and 3 h after necessary in two procedure added to patients (group not orange juice+0·45% mentioned) saline up to 12 h Overall CN before and after incidence 24% 400 mg twice daily SCrea 5 Chronic dialysis the day before and on (CAPD) needed in one the procedure day+ patient from PLA 0·45% saline 12 h group (0·8%) during before and after 3 months’ follow-up Overall CN incidence 13% 600 mg twice daily SCrea 25% No dialysis necessary the day before and on Length of hospital procedure day+ stay significantly 0·9% saline 12 h reduced (3·9 days before and 6 h after control vs 3·4 days, NAC p=0·02) Overall CN incidence 8%

NA=data not available. NAC=acetylcysteine. CN=contrast nephropathy. SCrea,=serum creatinine. ATA=angiotensin receptor II antagonist. ACEI=angiotensin converting enzyme inhibitor. RCM=radio contrast media. DM=diabetes mellitus. CHF=chronic heart failure. EF=ejection fraction. Coro=coronary angiography. PTCA=percutaneous transluminal coronary angioplasty. PTA=percutaneous transluminal angioplasty. periph=peripheral. preop=preoperative. PLA=placebo. All RCM administered intraarterially unless otherwise stated. *Data are mean (SD) or number of patients (%). SCrea units mg/L. †Administered orally+hydration 1 mL/kg/h intravenously before and after radiocontrast media as indicated. ‡Change (mg/L) or % from baseline 48 h after procedure.

Table 2: Characteristics of the trials eligible for meta-analysis

these studies used a quasi-randomised design per se34 or with respect to chronic renal insufficiency,31 used either historical35 or no placebo control group,32 or contained redundant data,33 and were therefore not eligible (table 1). Thus, seven studies14–20 fulfilled all inclusion criteria, and the agreement between the two reviewers for the eligibility of the relevant articles was 100%. Quality scoring of these trials according to the Jadad scale was done by the same authors and agreement was 100% (table 2). The median score was 3 (IQR 1–4). All trials included patients with chronic renal insufficiency with a mean serum creatinine of 123·76– 247·52 mol/L and used standardised periprocedural hydration regimens. Six15–20 of the seven studies included patients scheduled for cardiovascular interventions necessitating intra-arterial administration of contrast dye,

600

whereas one14 investigated intravenous administration for elective CT examinations. The mean amount of radiocontrast media given ranged from 75 mL to 187 mL, and all of them were non-ionic. The overall incidence of contrast nephropathy varied between 8% and 28%. Acetylcysteine was administered orally periprocedurally by slightly different regimens and formulations. 403 patients were randomised to acetylcysteine combined with intravenous hydration and 402 to intravenous hydration alone. Their mean age was 69 years (SD 3), the proportion of men ranged from 57% to 90%, and the number of patients with diabetes varied between 21% and 64%. Details of the characteristics of these seven trials are in table 2. The outcome measure of contrast nephropathy was reported in all trials including a total of 805 patients. Four

THE LANCET • Vol 362 • August 23, 2003 • www.thelancet.com

For personal use. Only reproduce with permission from The Lancet.

ARTICLES

Study

Risk ratio (95% CI)

% Weight

Study

Risk ratio (95% CI)

% Weight

Tepel14 Allaqaband19 Briguori16 Diaz-Sandoval17 Durham15 Shyu15 Kay19

0·11 (0·02–0·86) 1·19 (0·45–3·12) 0·59 (0·23–1·57) 0·18 (0·04–0·72) 1·20 (0·55–2·63) 0·14 (0·03–0·57) 0·32 (0·11–0·96)

8·1 16·7 16·7 12·5 18·7 12·1 15·3

Allaqaband18 Briguori16 Diaz-Sandoval17 Durham14 Shyu15 Kay19

1·19 (0·45–3·12) 0·59 (0·23–1·57) 0·18 (0·04–0·72) 1·20 (0·55–2·63) 0·14 (0·03–0·57) 0·32 (0·11–0·96)

18·2 18·2 13·4 20·6 13·0 16·6

Overall (95% CI)

0·44 (0·22–0·88)

Overall (95% CI)

0·49 (0·24–1·00)

0·01

0·1

1·0

10·0

0·01

Risk ratio Risk ratio (95% CI)

% Weight

Tepel14 Allaqaband19 Briguori16 Diaz-Sandoval17 Durham15 Shyu15 Kay19

4·2 18·3 18·2 8·9 27·8 8·4 14·2 0·56 (0·37–0·84)

Overall (95% CI) 0·1

1·0

10·0

Risk ratio

Study

0·01

0·1

1·0

10·0

Risk ratio (95% CI)

Study

Allaqaband18 Briguori16 Diaz-Sandoval17 Durham14 Shyu15 Kay19

19·1 19·0 9·3 29·0 8·7 14·9 0·60 (0·39–0·91)

Overall (95% CI) 0·01

Risk ratio

% Weight

0·1

1·0

10·0

Risk ratio

Figure 1: Relative risk of occurrence of contrast nephropathy for all seven trials

Figure 2: Relative risk of occurrence of contrast nephropathy after exclusion of Tepel trial

Data are calculated by either a random-effects model (upper) or a fixedeffects model (lower). Boxes are relative risk, lines are 95% CI. Areas of boxes are proportional to the respective study weight within the corresponding pooled analysis (see also weight values on the right). The pooled treatment estimate and its 95% CI are shown as a diamond with a dotted vertical line indicating the pooled estimate value at the bottom of each plot. The vertical solid line indicates no treatment effect; values of less than 1 indicate treatment benefit.

Data are calculated by either a random-effects model (upper) or a fixedeffects model (lower). Boxes are relative risk, lines are 95% CI. Areas of boxes are proportional to the respective study weight within the corresponding pooled analysis (see also weight values on the right). The pooled treatment estimate and its 95% CI are shown as a diamond with a dotted vertical line indicating the pooled estimate value at the bottom of each plot. The vertical solid line indicates no treatment effect; values of less than 1 indicate treatment benefit.

studies14,16,18,20 showed a significant reduction of the relative risk for development of contrast nephropathy in patients given acetylcysteine, whereas the remaining three showed no significant benefit.15,17,19 The Q statistic indicated significant heterogeneity between the trials (p=0·016). 8% (33 of 403) of patients given acetylcysteine and 18% (74 of 402) of controls developed contrast nephropathy. When the effect sizes of these seven trials were combined by a random-effects model a significant relative risk reduction of 56% was seen (p=0·02; figure 1). Table 3 shows the individual absolute and relative risks in each trial.

For sensitivity analysis, calculation was also done with a fixed-effects model for combining effect sizes of all studies with and without quality scoring and both random-effects and fixed-effects models after excluding the Tepel trial. With a fixed-effects model on all trials we recorded a significant reduction in relative risk (p=0·0054; figure 1), which was confirmed after taking into account a quality weight derived from the product of the inverse variance and the quality score (0·55 [0·35–0·87], p=0·010). After exclusion of the Tepel trial, we recorded a 51% (p=0·051) reduction in relative risk with the random-effects model and a 40% (p=0·016) reduction with the fixed-effects model (figure 2). By meta-regression, neither the volume of radiocontrast media administered nor the amount of chronic renal insufficiency was associated with the relative risk of contrast nephropathy (chronic renal insufficiency slope coefficient –0·95 [–3.35 to 1·45], p=0·44; radiocontrast media volume coefficient –0·01 [–0·04 to 0·02], p=0·45). Funnel plotting of the SE of the log relative risk versus the log relative risk showed assymmetry (figure 3), which was confirmed by formal statistical testing suggesting the presence of publication bias, particularly the absence of small negative trials (p for bias=0·001, Begg’s test or p=0·007, Egger’s test).

Trial Tepel14 Allaqaband18 Briguori16 Diaz-Sandoval17 Durham14 Shyu15 Kay19

NAC-treated group

Control group

Relative risk (95% CI)

1/41 (2%) 8/45 (18%) 6/92 (7%) 2/25 (8%) 10/38 (26%) 2/60 (3%) 4/102 (4%)

9/42 (21%) 6/40 (15%) 10/91 (11%) 13/29 (45%) 9/41 (22%) 15/61 (25%) 12/98 (12%)

0·11 (0·02–0·86) 1·18 (0·45–3·12) 0·59 (0·22–1·57) 0·18 (0·04–0·72) 1·20 (0·55–2·63) 0·14 (0·03–0·57) 0·32 (0·11–0·96)

Values are number of patients with contrast nephropathy/number of individuals in trial (%).

Table 3: Individual absolute and relative risks for development of contrast nephropathy

THE LANCET • Vol 362 • August 23, 2003 • www.thelancet.com

For personal use. Only reproduce with permission from The Lancet.

601

ARTICLES

2

LogRR

0

–2

–4 0

0·5

1·0

SE of logRR Figure 3: Assessment of publication bias RR=relative risk. Begg’s funnel plot: the horizontal line indicates the fixedeffects pooled estimate, the other lines are pseudo 95% CI. Asymmetry is present as indicated by the absence of data in the upper right side of the plot.

Discussion Contrast nephropathy is the third leading cause of acute renal failure in patients who had been admitted, accounting for 10% of all cases.37 Generally, it follows a benign course, and persistent renal impairment and dialysis dependency are rare.4 In selected subgroups of patients, however, like those with pre-existing renal insufficiency or diabetes mellitus, up to 7% require transient dialysis or progress to end-stage renal disease.5 The occurrence of contrast nephropathy has also been associated with increased mortality both in hospital and in the long-term, implicating an extended hospital stay and increased health care costs.5,6 Acetylcysteine, a thiol-containing antioxidant, is a promising agent to prevent contrast nephropathy because of its antioxidative and haemodynamic effects in the renal medulla and its general organ-protective effects described in several ischaemia-reperfusion models.12 Tepel and colleagues14 reported in 2000 an almost 90% relative risk reduction in the incidence of contrast nephropathy in patients with chronic renal insufficiency given acetylcysteine after intravenous administration of radiocontrast media for elective CT examinations. This finding provided the rationale to investigate use of acetylcysteine for prevention of contrast nephropathy in patients with chronic renal insufficiency undergoing cardiovascular procedures and receiving intra-arterial radiocontrast media. Since 2000, six trials14–19 have been published, with contrasting results. In our meta-analysis combining the effect sizes of all seven randomised controlled trials that used acetylcysteine for prevention of contrast nephropathy in chronic renal insufficiency through a random-effects model, we recorded a significant 56% relative risk reduction in patients given acetylcysteine. Our main result was robust to different assumptions during sensitivity analysis. Thus, in the fixed-effects model with and without quality scoring to combine the effect sizes of all trials and in the re-calculation after restricting our analysis to patients receiving radiocontrast media for cardiovascular interventions by excluding the Tepel study, comparable reductions in the relative risk of contrast nephropathy were obtained. Which statistical model is most appropriate to combine effect sizes of different trials is uncertain.23,26 However, if statistical heterogeneity is present between trials, as in our study, use of random-combined models is generally recommended.23,26 By contrast with fixed-effects models,

602

which calculate a weighted average effect under the assumption that outcome differences result from chance alone, random-effects models take into account that some outcomes might result from chance and between-study differences. Obtaining significance is therefore more difficult in random-effects models than fixed-effects models and the results are thus thought to be more generalisable.23,26,38 Some baseline characteristics of included patients and some study design details differed between the analysed trials (table 2), particularly the degree of chronic renal insufficiency before the procedure and the amount of radiocontrast media given. Since both variables have been reported as independent predictors of contrast nephropathy4,5,7 these discrepancies might have contributed to the heterogeneous results. However, metaregression showed no significant relation between these covariates and the relative risk of contrast nephropathy as a dependent variable. Our meta-analysis has several limitations that should be taken into account. First, the asymmetrical appearance of the funnel plot suggests that publication bias is present, particularly the absence of small trials with negative results. Thus, and despite doing a thorough search including not only several international but also a German database to reduce language bias, and manually searching the proceedings of major cardiology and nephrology meetings, we cannot rule out that publication bias might lead to an overestimation of the true treatment effect. Although such bias is known to be present in the published work,39 the asymmetry of the funnel plot, does not prove publication bias and might also result from, for example, differences in the underlying risk of contrast nephropathy. Second, all included studies used the surrogate endpoint of contrast nephropathy as a primary outcome. Contrast nephropathy was defined as an increase of serum creatinine of more than 44·2 mol/L or 25% from baseline values, which is, especially in patients with pre-existing renal insufficiency, a minor deterioration of renal function. Even in the high-risk cohorts included, contrast nephropathy was almost always transient and only rarely needed dialysis—indicated by an overall incidence of dialysis dependency of 0·7% (seven of 805; table 2). Despite the reported association of contrast nephropathy with impaired outcomes, particularly in patients with advanced renal failure or requiring dialysis after administration of radiocontrast media,4–7 no trial was designed to investigate the effect of acetylcysteine on hard clinical endpoints such as in-hospital morbidity, mortality, or dialysis dependency. Only the trial implemented by Kay and colleagues20 also investigated the effect of acetylcysteine on the secondary endpoint of length of hospital stay and found a significant reduction of a half day in patients given this drug. Thus, the clinical relevance of the renoprotective effect of acetylcysteine remains somewhat debatable40 whereas periprocedural use of adequate hydration regimens is of proven benefit in preventing contrast nephropathy as emphasised by Trivedi and colleagues.9 Our meta-analysis showed a significant benefit of acetylcysteine treatment in prevention of contrast nephropathy in patients with chronic renal insufficiency. Whether the observed reduction in relative risk of an arbitrarily defined increase in serum creatinine will confer benefit in clinical practice remains controversial. Nevertheless, the reported association of contrast nephropathy with increased morbidity, mortality, and lengthened hospital stay might justify use of acetylcysteine for prophylaxis of contrast nephropathy since it is cheap,

THE LANCET • Vol 362 • August 23, 2003 • www.thelancet.com

For personal use. Only reproduce with permission from The Lancet.

ARTICLES

easy to use, and has a favourable side-effect profile. Trials designed to investigate the course of serum creatinine after radiocontrast media application and to elucidate the effect of acetylcysteine on hard clinical endpoints are warranted. Contributors R Birck contributed to protocol design, data extraction, quality assessment, statistical analysis, and writing and revision of the report. S Krzossok helped with protocol design and writing and revision of the report. F Markowetz contributed to statistical analysis. P Schnuelle contributed to statistical analysis and revision of the report. F J van der Woude contributed to protocol design, interpretation of data, and revision of the report. C Braun contributed to protocol design, data extraction, quality assessment, and to writing and revision of the report.

Conflict of interest statement None declared.

Acknowledgments We thank S Allaqaband, L Diaz-Sandoval, K Shyu, and M Tepel for providing unpublished details from their individual trials.

References 1 2

3

4

5

6 7

8

9

10

11

12 13

14

15

16

Murphy SW, Barrett BJ, Parfrey PS. Contrast nephropathy. J Am Soc Nephrol 2000; 11: 177–82. Parfrey PS, Griffiths SM, Barrett BJ, et al. Contrast material-induced renal failure in patients with diabetes mellitus, renal insufficiency, or both. A prospective controlled study. N Engl J Med 1989; 320: 143–49. Rich MW, Crecelius CA. Incidence, risk factors, and clinical course of acute renal insufficiency after cardiac catheterization in patients 70 years of age or older: a prospective study. Arch Intern Med 1990; 150: 1237–42. McCullough PA, Wolyn R, Rocher LL, Levin RN, O’Neill WW. Acute renal failure after coronary intervention: incidence, risk factors, and relationship to mortality. Am J Med 1997; 103: 368–75. Gruberg L, Mintz GS, Mehran R, et al. The prognostic implications of further renal function deterioration within 48 h of interventional coronary procedures in patients with pre-existent chronic renal insufficiency. J Am Coll Cardiol 2000; 36: 1542–48. Levy EM, Viscoli CM, Horwitz RI. The effect of acute renal failure on mortality: a cohort analysis. JAMA 1996; 275: 1489–94. Rihal CS, Textor SC, Grill DE, et al. Incidence and prognostic importance of acute renal failure after percutaneous coronary intervention. Circulation 2002; 105: 2259–64. Solomon R, Werner C, Mann D, D’Elia J, Silva P. Effects of saline, mannitol, and furosemide to prevent acute decreases in renal function induced by radiocontrast agents. N Engl J Med 1994; 331: 1416–20. Trivedi HL, Moore H, Nasr S, et al. A randomised prospective trial to assess the role of saline hydration on the development of contrast nephropathy. Nephron Clin Pract 2003; 93: c29–34. Bakris GL LN, Gaber AO, Jones JD, Burnett JC. Radiocontrast medium-induced declines in renal function: a role for oxygen radicals. Am J Physiol 1990; 258: F115–20. Yoshioka T, Fogo A, Beckman JK. Reduced activity of antioxidant enzymes underlies contrast media-induced renal injury in volume depletion. Kidney Int 1992; 41: 1008–15. Tepel M, Zidek W. Acetylcysteine and contrast media nephropathy. Curr Opin Nephrol Hypertens 2002; 11: 503–06. Safirstein R, Andrade L, Vieira JM. Acetylcysteine and nephrotoxic effects of radiographic contrast agents: a new use for an old drug. N Engl J Med 2000; 343: 210–12. Tepel M, van der Giet M, Schwarzfeld C, Laufer U, Liermann D, Zidek W. Prevention of radiographic-contrast-agent-induced reductions in renal function by acetylcysteine. N Engl J Med 2000; 343: 180–84. Durham JD, Caputo C, Dokko J, et al. A randomized controlled trial of N-acetylcysteine to prevent contrast nephropathy in cardiac angiography. Kidney Int 2002; 62: 2202–07. Shyu KG, Cheng JJ, Kuan P. Acetylcysteine protects against acute renal damage in patients with abnormal renal function undergoing a coronary procedure. J Am Coll Cardiol 2002; 40: 1383–88.

17 Briguori C, Manganelli F, Scarpato P, et al. Acetylcysteine and contrast agent-associated nephrotoxicity. J Am Coll Cardiol 2002; 40: 298–303. 18 Diaz-Sandoval LJ, Kosowsky BD, Losordo DW. Acetylcysteine to prevent angiography-related renal tissue injury (the APART trial). Am J Cardiol 2002; 89: 356–58. 19 Allaqaband S, Tumuluri R, Malik AM, et al. Prospective randomized study of N-acetylcysteine, fenoldopam, and saline for prevention of radiocontrast-induced nephropathy. Catheter Cardiovasc Interv 2002; 57: 279–83. 20 Kay JCW, Chan TM, Lo SK, et al. Acetylcysteine for prevention of acute deterioration of renal function following elective coronary angiography and intervention. JAMA 2003; 289: 553–58. 21 Moher D CD, Eastwood S, Olkin I, Rennie D, Stroup DF, for the QUOROM Group. Improving the quality of reports of meta-analysis of randomised controlled trials: the QUOROM statement. Lancet 1999; 354: 1896–900. 22 Jadad ARMA, Caroll D, Jenkinson C, Reynolds DJMR, Gavaghan DJ, McQuay J. Assessing the quality of reports of randomized clinical trials: Is blinding necessary? Control Clin Trials 1996; 17: 1–12. 23 Lau J, Ioannidis, JPA, Schmid CH. Quantitative synthesis in systematic reviews. Ann Intern Med 1997; 127: 820–26. 24 DerSimonian RLN. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–88. 25 Thompson SGSS. Explaining heterogeneity in meta-analysis: a comparison of methods. Stat Med 1999; 18: 2693–708. 26 Egger M, Smith GD, Phillips AN. Meta-analysis: principles and procedures. BMJ 1997; 315: 1533–37. 27 Moher DPB, Jones A, Cook DJ, et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analysis? Lancet 1998; 352: 609–13. 28 Sterne JACEM, Smith GD. Investigating and dealing with publication and other biases in meta-analysis. BMJ 2001; 323: 101–05. 29 Begg CBMM. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994; 50: 1088–101. 30 Egger MSG, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ; 315: 629–34. 31 Vallero A, Cesano G, Pozzato M, et al. Contrast nephropathy in cardiac procedures: no advantages with prophylactic use of Nacetylcysteine (NAC). G Ital Nefrol 2002; 19: 529–33. 32 Bader BD, Bergen ED, Rossmann S, et al. Prevention of contrast media-induced acute renal failure in patients with severe impaired renal function: theophylline versus N-acetylcysteine. J Am Soc Nephrol 2002; 13: 447A. 33 Tepel M, van der Giet M, Schwarzfeld C, Laufer U, Liermann D, Zidek W. Acetylcysteine ameliorates radiographic-contrast-agentinduced reductions in renal function in diabetic patients with chronic renal failure. Nephrol Dial Transplant 2001; 16: A53. 34 Adamian MG, Moussa I, Mehran R, et al. The role of Mucomyst administration prior to percutaneous interventions on renal function in patients with chronic renal failure. J Am Coll Cardiol 2002, 39 (suppl): 5A. 35 Mouhayar EN, Tadros G, Akinwande A, Campbell B, Blankenship JC, Illiadis EA. Prevention of contrast-induced renal dysfunction with acetylcysteine in patients undergoing coronary angiography. J Am Coll Cardiol 2002; 39 (suppl): 5A. 36 van der Giet M, Ziedel W, Tepel M. Deterioration of renal function in diabetic patients with chronic renal failure by radiographiccontrast-agent can be attenuated by acetylcysteine. J Am Soc Nephrol 2001; 12: 849A. 37 Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT. Hospital-acquired renal insufficiency: a prospective study. Am J Med 1983; 74: 243–48. 38 Furukawa T, Guyatt GH, Grifitth LE. Can we individualize the number needed to treat (NNT)? An empirical study of summary effect measures in meta-analyses. Int J Epidemiol 2002; 31: 72–76. 39 Sterne JAC, Egger M, Smith GD. Investigating and dealing with publication and other biases. In: Egger MD, Smith GD, Altman DG, eds. Systematic reviews in health care: meta-analysis in context. London: BMJ Books, 2001. 40 Conti CR. Contrast nephropathy or contrast creatininopathy. Clin Cardiol 2003; 26: 53–54.

THE LANCET • Vol 362 • August 23, 2003 • www.thelancet.com

For personal use. Only reproduce with permission from The Lancet.

603