Emottawa.ca

Annals of Internal Medicine
Meta-analysis: Effectiveness of Drugs for Preventing Contrast-Induced
Nephropathy
Aine M. Kelly, MD, MS; Ben Dwamena, MD; Paul Cronin, MD, MS; Steven J. Bernstein, MD, MPH; and Ruth C. Carlos, MD, MS
Background: N-Acetylcysteine, theophylline, and other agents have
Data Synthesis: In the 41 studies included, N-acetylcysteine (rela-
shown inconsistent results in reducing contrast-induced nephropa- tive risk, 0.62 [95% CI, 0.44 to 0.88]) and theophylline (relative risk, 0.49 [CI, 0.23 to 1.06]) reduced the risk for contrast-inducednephropathy more than saline alone, whereas furosemide increased Purpose: To determine the effect of these agents on preventing
it (relative risk, 3.27 [CI, 1.48 to 7.26]). The remaining agents did not significantly affect risk. Significant subgroup heterogeneity was Data Sources: Relevant randomized, controlled trials were identi-
present only for N-acetylcysteine. No publication bias was dis- fied by computerized searches in MEDLINE (from 1966 through 3 November 2006), EMBASE (1980 through November 2006), Limitations: All trials evaluated the surrogate end point of contrast-
PubMed, Web of Knowledge (Current Contents Connect, Web ofScience, BIOSIS Previews, and ISI Proceedings for the latest 5 induced nephropathy as the primary outcome. The lack of a sta- years), and the Cochrane Library databases (up to November tistically significant renoprotective effect of theophylline may result 2006). Databases were searched for studies in English, Spanish, from insufficient data or study heterogeneity. True study quality Study Selection: Randomized, controlled trials that administered
Conclusion: N-Acetylcysteine is more renoprotective than hydra-
N-acetylcysteine, theophylline, fenoldopam, dopamine, iloprost, sta- tion alone. Theophylline may also reduce risk for contrast-induced tin, furosemide, or mannitol to a treatment group; used intravenous nephropathy, although the detected association was not significant.
iodinated contrast; defined contrast-induced nephropathy explicitly; Our data support the administration of N-acetylcysteine prophy- and reported sufficient data to construct a 2 ϫ 2 table of the laxis, particularly in high-risk patients, given its low cost, availability, Data Extraction: Abstracted information included patient charac-
teristics, type of contrast media and dose, periprocedural hydration,
definition of contrast-induced nephropathy, and prophylactic agent
Ann Intern Med. 2008;148:284-294.
www.annals.org
For author affiliations, see end of text.
Contrast-induced nephropathy, defined as an increase osmolarcontrastagentshasbeenshowntodecreasetherisk
in serum creatinine greater than 25% or 44.2 ␮mol/L for contrast-induced nephropathy (11-13). These mea- (Ͼ0.5 mg/dL) within 3 days of intravascular contrast ad- sures suffice for many patients; however, the risk is reduced ministration in the absence of an alternative cause, is the but not eliminated in some patients- even when iso- third most common cause of new acute renal failure in osmolar contrast is used (14, 15). Other studies have eval- hospitalized patients (1, 2). Contrast-induced nephropathy uated the use of N-acetylcysteine, theophylline, fenoldo- develops in 0% to 10% of patients with normal renal func- pam, and other agents as preventive strategies in contrast- tion (3). However, the incidence may be as high as 25% in induced nephropathy; the results have been heterogeneous patients with preexisting renal impairment or certain risk and are difficult to compare across the different treatment factors, such as diabetes, congestive heart failure, advanced strategies. Given the widespread use of iodinated intravas- age, and concurrent administration of nephrotoxic drugs cular contrast agents, an improved understanding of the (3). Large doses of intravenous contrast and use of high- potential value of these agents has important patient safety osmolar contrast agents in patients with renal impairment also increase the risk for contrast-induced nephropathy (4 -6). High-osmolar contrast agents are more rarely used now.
The risk difference between iso-osmolar agents, such as iodixanol, and low-osmolar agents, such as iopamidol, ioxaglate, or iohexol, is less clear (7-9). Most episodes of Editors' Notes. 285 contrast-induced nephropathy are not detected clinically be-cause patients are asymptomatic. However, contrast- Web-Only
induced nephropathy may increase the risk for renal failure and is associated with dialysis, prolonged hospital stay, in- creased health care costs, potentially irreversible reduction Use of preprocedural fluids and low-osmolar or iso- 284 19 February 2008 Annals of Internal Medicine Volume 148 • Number 4
www.annals.org
Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy Review We conducted a meta-analysis of the literature to quantify the effects of individual strategies on the preven- tion of contrast-induced nephropathy and to facilitate Contrast-induced nephropathy is a common cause of comparison of preventive effects across strategies.
acute renal failure in hospitalized patients. Clinicians use avariety of contrast agents to reduce the risk for contrast-induced nephropathy, including N-acetylcysteine, theo- phylline, fenoldopam, dopamine, furosemide, mannitol, Study Search Strategy
We performed a computerized search by using stan- Contribution
dard meta-analytic techniques (16) to identify relevant ar-ticles in MEDLINE (from 1966 through 3 November Although all of the agents included in this analysis reducedthe risk for contrast-induced nephropathy, this meta-anal- 2006), EMBASE (1980 through November 2006), ysis of 33 trials involving 3622 patients found the stron- PubMed, Web of Knowledge (Current Contents Connect, gest evidence for the effectiveness of N-acetylcysteine, Web of Science, BIOSIS Previews, and ISI Proceedings for mannitol, and theophylline when compared with peri- the latest 5 years), and the Cochrane Library databases. For the MEDLINE search, we used the following combinationof keywords: [renal failure or kidney failure to include all subheadings] and [contrast media or iopamidol or iodine Available studies examined laboratory end points (such as or ioxaglic acid or iodine compounds or iohexol or urog- an increase in serum creatinine levels) rather than clinical raphy or drug hyper sensitivity or tomography, X ray com- end points (such as dialysis or death).
puted or diatrizoate] and [hydration or fluid therapy orwater or dehydration or skin or nutritional support or body water] and [clinical trial or randomized controlled trial]and [prospective trial or prospective studies or clinical tri-als] and [adult or middle aged or aged] and [N-Acetylcys- mellitus or hypertension, and mean baseline creatinine teine or acetylcysteine] or [theophylline] or [mannitol] or level), type of radiologic or cardiologic imaging, inclusion [dopamine] or [fenoldopam] or [bicarbonate]. For the and exclusion criteria, type of contrast media and dose PubMed, Cochrane Library Database, and Web of Knowl- used, periprocedural hydration, specific definition of con- edge searches, we used the search words renal failure, con- trast-induced nephropathy, prophylactic agent dose and trast medium, hydration, randomized controlled trial, N route, and serum creatinine level at baseline and at 48 acetyl cysteine, Theophylline, Mannitol, Fenoldopam, Dopa- mine and Bicarbonate. We included English-, French-, Analysis of Renoprotective Agents
German-, Spanish- and Italian-language studies and clini- The primary outcome was the development of con- cal trials and excluded review articles and nonhuman stud- trast-induced nephropathy, defined as an absolute increase ies. We combined this strategy with a manual search of in baseline serum creatinine greater than 44.2 ␮mol/L reference lists from identified articles.
(Ͼ0.5 mg/dL) or a relative increase greater than 25% at 48hours after contrast injection. For trials missing this da- Study Selection
tum, we contacted the original authors to get the number We included a study if 1 of the treatment groups re- of patients with this outcome. We calculated individual ceived N-acetylcysteine, theophylline, fenoldopam, ilo- study relative risks and 95% CIs before aggregation. Sub- prost, statin, dopamine, trimetazidine, bicarbonate, ascor- sequently, we obtained overall and subgroup summary risk bic acid, furosemide, or mannitol. Criteria for inclusion ratios by random-effects modeling of the binary data from were randomized, controlled trials that compared treat- the multiple 2 ϫ 2 tables. We used the method of Der- ment with control; used intravenous iodinated contrast; Simonian and Laird (17), with the estimate of heterogene- explicitly defined contrast-induced nephropathy; and suffi- ity taken from the inverse variance fixed-effect model. We ciently reported data to construct a 2 ϫ 2 table and calcu- used the metan module in Stata, version 9.0 (Stata, College late the primary effect measure (relative risk reduction).
Station, Texas), to perform data synthesis.
Where data were missing, we contacted the original au- We performed subgroup evaluation of each therapeu- tic regimen. In studies comparing 2 dosage regimens of the Data Extraction
same intervention with a single control group (18 -20), we One reviewer examined the abstracts to determine considered the same-study dosage groups as representing a whether the study met the inclusion and exclusion criteria.
single intervention to avoid double-counting of shared Two reviewers separately abstracted complete articles ac- control observations. When we identified only 1 study that cording to a standardized form for studies meeting criteria.
examined a given therapy, we assigned that study to a Abstracted information included patient characteristics group termed "other" and pooled data from all such stud- (mean age, proportion of men and patients with diabetes ies together. This group included 1 study each on the use www.annals.org
19 February 2008 Annals of Internal Medicine Volume 148 • Number 4 285
Review Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy effect estimates that is due to genuine variation rather than Figure 1. Trial identification, inclusion, and exclusion.
sampling error (22). A value of 0% indicates no observedheterogeneity. Higgins and colleagues (22) suggest describ- Abstracts retrieved for detailed
ing I2 values of 25%, 50%, and 75% as low, moderate, evaluation (n = 88)
and high, respectively. We obtained the group-specificand overall I2 as standard output of the metan program.
Excluded reviews (n = 23):
Nonhuman trials
We performed an Egger precision-weighted linear re- Nonclinical outcomes
gression test as a statistical test of funnel plot asymmetry Language other than
English, French, German,
Spanish, or Italian
All statistical analyses were performed with Stata.
Full-text articles retrieved for
detailed evaluation (n = 65)
RESULTS
Study Identification
Excluded articles (n = 24):
Our initial search yielded 619 citations and references.
Not English, French,
German, Spanish, or
We excluded 531 studies on the basis of our criteria, in- cluding nonclinical trials; trials not conducted on humans; Nonhuman trials
Nonrandomized clinical
trials not reported in English, French, German, Spanish, or Italian; trials reporting only nonnephropathy outcomes; Nonnephropathy outcome
and trials using nonclinical outcome measures, leaving 88 Articles describing the
studies that met the inclusion criteria (Figure 1). We re-
following treatment regimens
viewed abstracts from the 88 articles and excluded an ad- N-Acetylcysteine: 30
ditional 23 trials, including nonrandomized clinical trials; Theophylline: 6
trials not conducted on humans; trials not reported in En- Dopamine: 3
Fenoldopam: 2
glish, French, German, Spanish, or Italian; trials reporting Furosemide: 2
only nonnephropathy outcomes; and trials that used non- Iloprost: 2
Statin: 1
clinical outcome measures, leaving 65 studies for full pub- Ascorbic acid: 1
lication review. The full articles were then reviewed, and a Bicarbonate: 1
Furosemide and dopamine: 1
further 24 studies were excluded for reasons similar to Furosemide, dopamine, and
those just mentioned. After the final screening, 41 random- mannitol: 1
Mannitol: 1
ized clinical trials met our inclusion criteria (18 -20, 24 -59), involving 6379 patients who had elective radiographicprocedures involving contrast agents.
of iloprost; trimetazidine; mannitol; bicarbonate; ascorbic Study Characteristics
acid; and combinations of furosemide, dopamine, and The trials were published between 1994 and 2006, mannitol and furosemide and dopamine. We used relative and the Table shows their characteristics. Fifteen trials
risk ratios to estimate the treatment effects.
were performed in the United States (9, 20, 24, 25, 30 -32, 34, 44, 48, 49, 51, 52, 58, 59), and 26 trials were Assessment of Methodological Quality
performed elsewhere (18, 19, 26 -29, 33, 35- 43, 45- 47, Criteria for quality assessment included concealment 50, 53-57, 59). Thirty-four trials evaluated patients with of allocation, similarity of both groups at baseline regard- impaired renal function (9, 19, 20, 24 -29, 32- 46, 48 -52, ing prognostic indicators, eligibility criteria, blinding of 56 -59), defined as serum creatinine levels greater than patient, blinding of care provider, blinding of outcome 106.1 to 132.6 ␮mol/L (Ͼ1.2 to 1.5 mg/dL). We had assessor, point estimates and measures of variability for the insufficient data to separately evaluate patients with normal primary outcome measure, and inclusion of an intention- renal function. Only 3 trials evaluated patients with nor- to-treat analysis (21). Any disagreements in abstracted data mal and impaired renal function, and 2 trials evaluated between the reviewers were adjudicated by a third reviewer.
only patients with normal renal function (30, 31, 47, 53, We explored potential heterogeneity in estimates of treat- 54). One trial evaluated patients having computed tomog- ment efficacy attributable to each quality criterion by using raphy (58); the rest evaluated patients having cardiac cath- eterization. The average age of the study patients was Assessment of Heterogeneity
greater than 65 years in all but 8 studies (9, 18, 29, 37, 40, We used Forest plots to visualize the extent of hetero- 41, 45, 48), and all studies included patients with diabetes.
geneity among studies. We also examined I2, a standard Dosing regimens for each trial are detailed in the Table.
test for heterogeneity that measures the degree of inconsis- The outcome measure of contrast-induced nephropathy tency across studies. I2 values, which range from 0% to was reported in all studies. Changes in serum creatinine 100%, describe the proportion of variation in treatment levels were reported at 48 hours in most trials (9, 18 -20, 286 19 February 2008 Annals of Internal Medicine Volume 148 • Number 4
www.annals.org
Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy Review Table. Study Characteristics*
Study, Year
Patients in the
Treatment
Patients in
Enrollment Criteria
Hydration Protocol
(Reference)
Patients,
Prophylactic Agent
Saline-
Only
Group, n
Dopamine
Fenoldapam
per h with cardiacfailure), for 2 to12 h beforeallocation Furosemide
N-Acetylcysteine
www.annals.org
19 February 2008 Annals of Internal Medicine Volume 148 • Number 4 287
Review Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy Table-Continued
Study, Year
Patients in the
Treatment
Patients in
Enrollment Criteria
Hydration Protocol
(Reference)
Patients,
Prophylactic Agent
Saline-
Only
Group, n
per h, up to 12 hbefore and 12 hafter contrastadministration mg, before angioplasty,and 1200 mg twice dailyfor 48 h after mL, over 4 to 6 hfor 6 to 12 hbefore and afterangiography 288 19 February 2008 Annals of Internal Medicine Volume 148 • Number 4
www.annals.org
Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy Review Table-Continued
Study, Year
Patients in the
Treatment
Patients in
Enrollment Criteria
Hydration Protocol
(Reference)
Patients,
Prophylactic Agent
Saline-
Only
Group, n
between 0.13 and0.66 mL/s (8 and40 mL/min) and saline, 1.5mL/kg per h, for6 h or untildischarge Other‡
IV, 1 mL/kg perh, for 8 h beforeand after contrastin both groups Theophylline
www.annals.org
19 February 2008 Annals of Internal Medicine Volume 148 • Number 4 289
Review Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy Table-Continued
Study, Year
Patients in the
Treatment
Patients in
Enrollment Criteria
Hydration Protocol
(Reference)
Patients,
Prophylactic Agent
Saline-
Only
Group, n
* All studies evaluated nephropathy after IV angiography except where indicated. CAD ϭ coronary artery disease; Cr ϭ creatinine; CrC ϭ creatinine clearance; CRI ϭ chronic renal impairment; GFR ϭ glomerular filtration rate; IV ϭ intravenous; NaCl ϭ sodium chloride.
† Study evaluated computed tomography.
‡ This category comprised all studies that were the only included studies for a given therapy.
24 - 47, 49 -59), although we used outcomes reported at characteristic of explicitly stating specific inclusion criteria (P ϭ 0.007) independently contributed to heterogeneity Analysis of Renoprotective Agents
Of the evaluated agents, N-acetylcysteine significantly Assessment of Within-Group Heterogeneity and
decreased the risk for contrast-induced nephropathy com- Publication Bias
pared with saline alone (relative risk, 0.62 [95% CI, 0.44 Treatment effect estimates within the N-acetylcysteine to 0.88]) (Figure 2). Although seemingly renoprotective,
group showed moderate heterogeneity (I2 ϭ 55%; P Ͻ the effects of theophylline on nephropathy prevention were 0.001). As expected, we found a moderate to high level of not significant (relative risk, 0.49 [CI, 0.23 to 1.06]). In heterogeneity among the pooled studies that each exam- the heterogeneous group of treatments for which only a ined a different therapy (labeled "other") (I2 ϭ 61%; P ϭ single study was identified (labeled "other"), only ascorbic 0.024). No other groups demonstrated significant within- acid (relative risk, 0.46 [CI, 0.23 to 0.90]) and bicarbonate group heterogeneity (Figure 2). No significant publication
(relative risk, 0.12 [CI, 0.02 to 0.95]) significantly reduced bias was discerned (bias coefficient, Ϫ0.55; P ϭ 0.20).
contrast-induced nephropathy. Furosemide (relative risk,3.27 [CI, 1.48 to 7.26]) increased the risk for contrast-induced nephropathy.
DISCUSSION
Assessment of Methodological Quality
In our meta-analysis of 41 randomized trials, we found The Appendix Table (available at www.annals.org)
that preprocedural treatment with N-acetylcysteine effec- presents the quality characteristics of each study. Most tively reduced the risk for contrast-induced nephropathy.
studies included patients with similar baseline characteris- Theophylline also produced larger risk reductions than tics (94%) or specific inclusion characteristics (90%). Most previously mentioned; however, the effects of this agent also presented variance estimates of treatment effects were not significant. Not all agents analyzed had beneficial (59%) or blinding of patients to treatment (51%). Fewer effects-fenoldopam; furosemide; mannitol; and the com- than half of the studies reported concealment of allocation bination of furosemide, dopamine, and mannitol had odds (47%) or blinding of care providers to treatment (43%).
ratios greater than 1. Our findings for N-acetylcysteine Few studies noted outcome evaluation by individuals support previous studies (60 - 63). To date, no meta-anal- blinded to treatment assignment (6%) or an intention-to- yses have studied preprocedural dopamine or statins for the treat design (8%). In exploratory analysis, only the quality prevention of contrast-induced nephropathy. Our findings 290 19 February 2008 Annals of Internal Medicine Volume 148 • Number 4
www.annals.org
Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy Review Figure 2. Forest plot describing relative risk for contrast-induced nephropathy, by treatment agent.
Study, Year (Reference)
Relative Risk (95% CI)
Intervention, n/n Control, n/n
Dopamine
Abizaid et al., 1999 (24)
0.70 (0.33-1.47)
Diez et al., 1999 (31)
0.71 (0.26-1.95)
Subtotal (I2 = 0.0%; P = 0.98)
0.70 (0.39-1.28)
Fenoldopam
Allaqaband et al., 2002 (25)
1.05 (0.37-2.98)
Stone et al., 2003 (57)
1.11 (0.79-1.57)
Subtotal (I2 = 0.0%; P = 0.92)
1.11 (0.80-1.53)
Furosemide
Dussol et al., 2006 (35)
2.92 (0.99-8.67)
Solomon et al., 2006 (9)
3.73 (1.16-12.05)
Subtotal (I2 = 0.0%; P = 0.77)
3.27 (1.48-7.26)
N-Acetylcysteine
Allaqaband et al., 2002 (25)
1.19 (0.45-3.12)
Azmus et al., 2005 (26)
0.84 (0.43-1.67)
Baker et al., 2003 (27)
0.24 (0.05-1.05)
Balderramo et al., 2004 (28)
0.42 (0.04-4.44)
Briguori et al., 2002 (29)
0.59 (0.23-1.57)
Coyle et al., 2006 (30)
6.09 (0.75-49.24)
Diaz-Sandoval et al., 2002 (32)
0.18 (0.04-0.72)
Drager et al., 2004 (33)
0.42 (0.04-4.06)
Durham et al., 2002 (34)
1.20 (0.55-2.63)
El Mahmoud et al., 2003 (36)
1.50 (0.26-8.66)
Fung et al., 2004 (38)
1.30 (0.49-3.46)
Goldenberg et al., 2004 (39)
1.27 (0.30-5.31)
Gomes et al., 2005 (40)
1.03 (0.41-2.60)
Kay et al., 2003 (46)
0.32 (0.11-0.96)
Kefer et al., 2003 (47)
0.64 (0.11-3.68)
MacNeill et al., 2003 (48)
0.15 (0.02-1.11)
Marenzi et al., 2006 (18)
0.22 (0.13-0.37)
Namgung et al., 2005 (50)
0.37 (0.13-1.01)
Ochoa et al., 2004 (51)
0.33 (0.10-1.10)
Oldemeyer et al., 2003 (52)
1.28 (0.30-5.41)
Rashid et al., 2004 (53)
1.04 (0.22-4.91)
Sandhu et al., 2006 (54)
7.00 (0.37-132.29)
Shyu et al., 2002 (55)
0.14 (0.03-0.57)
Sinha et al., 2004 (56)
0.83 (0.28-2.48)
Tepel et al., 2000 (58)
0.11 (0.02-0.86)
Webb et al., 2004 (60)
1.07 (0.63-1.82)
Subtotal (I2 = 55.1%; P = 0.000)
0.62 (0.44-0.88)
Theophylline
Abizaid et al., 1999 (24)
0.60 (0.27-1.34)
Dussol et al., 2006 (35)
1.44 (0.42-4.92)
Erley et al., 1999 (37)
1.66 (0.16-17.37)
Huber et al., 2002 (42)
0.25 (0.06-1.12)
Huber et al., 2003 (43)
0.20 (0.05-0.87)
Kapoor et al., 2003 (45)
0.07 (0.00-1.12)
Subtotal (I2 = 39.7%; P = 0.141)
0.49 (0.23-1.06)
NOTE: Weights are from random-effects analysis
The intervention and control columns show the number of events among the total number of participants randomly assigned to the group for each study.
We estimated heterogeneity within subgroups by using the I2 statistic.
www.annals.org
19 February 2008 Annals of Internal Medicine Volume 148 • Number 4 291
Review Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy for theophylline support previous studies that showed a lished trials. The exclusion of unpublished data is generally risk reduction (64). However, the effects of theophylline associated with an overestimate of the true effect in meta- were not statistically significant in our study. In contrast, analysis (67). The single most common reason for inability Ix and colleagues (64) found borderline statistical signifi- to publish a trial is the lack of statistical significance, al- cance when they limited their analysis to studies using con- though some have suggested that the quality of unpub- comitant intravenous fluids or contrast volumes greater lished data is not comparable to that accepted by peer- than 100 mL and no statistical significance when their reviewed journals (68). In addition, many of the included analysis was limited to studies of only coronary angiogra- studies did not have high quality scores, and many did not phy patients or where the theophylline was given within 1 specify that they met the quality criteria, with the true hour of the procedure. Bagshaw and Ghali (65), however, did not find a statistically significant effect, similar to our Strengths of our study include the comprehensive search strategy and the careful statistical methods used. We N-Acetylcysteine is extremely inexpensive at 23 cents identified 41 trials with a total of 6379 patients and eval- for a 500-mg tablet (price as of 17 January 2007 at www uated multiple therapeutic agents within 1 analysis frame- .shopping.com), is readily available, and is easily adminis- work, allowing side-by-side comparison of the efficacies tered. Side effects and drug interactions are very rare with continued use and are highly unlikely to result from the Our meta-analysis shows that N-acetylcysteine is the limited use for renal protection. Therefore, although no most effective agent for preventing contrast-induced ne- formal cost-effective analysis has been performed to date, phropathy in patients with chronic renal insufficiency.
these findings support the use of N-acetylcysteine in se- Whether this risk reduction translates into a benefit in clinical outcomes remains to be proven. The reported as- Hydration and iso-osmolar or low-osmolar contrast sociation of contrast-induced nephropathy with increased agents, such as iodixanol, are all associated with a decreased morbidity, mortality, and hospital stay might justify the incidence of contrast-induced nephropathy in patients use of N-acetylcysteine as a routine intervention for pro- with renal impairment (creatinine clearance Ͻ1 mL/s phylaxis of contrast-induced nephropathy, given that [Ͻ60 mL/min]) (44). In a recent meta-analysis, McCul- N-acetylcysteine is readily available and inexpensive and lough and colleagues (15) found that although low-osmo- has a favorable side effect profile.
lar contrast agents reduced the risk for contrast-induced The results of this meta-analysis should be evaluated in nephropathy by two thirds, they did not totally eliminate head-to-head empirical studies of active agents to identify the risk. Thus, protective agents must still be considered the most efficacious regimen for preventing contrast- for patients with severe renal impairment who are to re- induced nephropathy. However, our findings indicate that ceive large volumes of contrast agents.
the use of such oral agents as N-acetylcysteine is reasonable Our meta-analysis has several limitations. All included in high-risk patients who are to receive large or repeated trials evaluated the surrogate end point of contrast-induced volumes of contrast agents. We believe that the lack of nephropathy as the primary outcome. Contrast-induced significant side effects and the low cost justifies use of these nephropathy was defined as an increase in serum creatinine agents while empirical data on clinical outcomes mature.
of more than 44.2 ␮mol/L (Ͼ0.5 mg/dL) or 25% frombaseline values, which represents a minor deterioration in From the University of Michigan and Veteran Affairs Ann Arbor Health- renal function in patients with chronic renal failure. Even in high-risk patients, contrast-induced nephropathy is al- Grant Support: Funded in part by the National Institutes of Health and
most always transient and only rarely requires dialysis.
the National Cancer Institute (grant 1 K07 CA108664 01A1) and the Only the trial by Kay and colleagues (46) examined length General Electric-Association of University Radiologists Radiology Re- of hospital stay as an end point and found a significant reduction in length of stay among patients given N-acetyl-cysteine. Despite the reported association of contrast- Potential Financial Conflicts of Interest: None disclosed.
induced nephropathy with impaired outcomes, no trial hasexamined clinical end points, such as dialysis dependency Requests for Single Reprints: Aine M. Kelly, MD, MS, Department of
or in-hospital morbidity and mortality. The clinical rele- Radiology, Division of Cardiothoracic Radiology, University of Michi- vance of the renoprotective effects of N-acetylcysteine, dopa- gan Hospitals, B1 132K Taubman Center, 1500 East Medical CenterDrive, Ann Arbor, MI 48109; e-mail, ainekell@med.umich.edu.
mine, and other agents is therefore debatable, whereasperiprocedural hydration is of proven benefit (66). In ad- Current author addresses are available at www.annals.org.
dition, it is possible that we did not detect a significanteffect for theophylline because of study heterogeneity orinsufficient data.
References
Because we primarily identified and used published 1. Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT. Hospital-
studies, our results are weighted on the findings of pub- acquired renal insufficiency: a prospective study. Am J Med. 1983;74:243-8.
292 19 February 2008 Annals of Internal Medicine Volume 148 • Number 4
www.annals.org
Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy Review after coronary angioplasty in patients with preexisting renal insufficiency. Am J 2. Morcos SK. Contrast media-induced nephrotoxicity- questions and answers.
Cardiol. 1999;83:260-3, A5. [PMID: 10073832] Br J Radiol. 1998;71:357-65. [PMID: 9659127] 25. Allaqaband S, Tumuluri R, Malik AM, Gupta A, Volkert P, Shalev Y, et al.
3. Morcos SK, Thomsen HS, Webb JA. Contrast-media-induced nephrotoxic-
Prospective randomized study of N-acetylcysteine, fenoldopam, and saline for ity: a consensus report. Contrast Media Safety Committee, European Society of prevention of radiocontrast-induced nephropathy. Catheter Cardiovasc Interv.
Urogenital Radiology (ESUR). Eur Radiol. 1999;9:1602-13. [PMID: 10525875] 4. Cigarroa RG, Lange RA, Williams RH, Bedotto JB, Hillis LD. Underesti-
26. Azmus AD, Gottschall C, Manica A, Manica J, Duro K, Frey M, et al.
mation of cardiac output by thermodilution in patients with tricuspid regurgita- Effectiveness of acetylcysteine in prevention of contrast nephropathy. J Invasive tion. Am J Med. 1989;86:417-20. [PMID: 2648822] 5. Murphy SW, Barrett BJ, Parfrey PS. Contrast nephropathy. J Am Soc Neph-
27. Baker CS, Wragg A, Kumar S, De Palma R, Baker LR, Knight CJ. A rapid
protocol for the prevention of contrast-induced renal dysfunction: the RAPPID 6. Thomsen HS, Morcos SK. Contrast-medium-induced nephropathy: is there a
study. J Am Coll Cardiol. 2003;41:2114-8. [PMID: 12821233] new consensus? A review of published guidelines. Eur Radiol. 2006;16:1835-40.
28. Balderramo DC, Verdu MB, Ramacciotti CF, Cremona LS, Lemos PA,
Ori'as M, et al. Renoprotective effect of high periprocedural doses of oral N-ace-
7. Morcos SK. Prevention of contrast media nephrotoxicity-the story so far.
tylcysteine in patients scheduled to undergo a same-day angiography. Rev Fac Clin Radiol. 2004;59:381-9. [PMID: 15081843] Cien Med Univ Nac Cordoba. 2004;61:13-9. [PMID: 16211988] 8. Thomsen HS, Morcos SK; ESUR. ESUR guidelines on contrast media. Ab-
29. Briguori C, Manganelli F, Scarpato P, Elia PP, Golia B, Riviezzo G, et al.
dom Imaging. 2006;31:131-40. [PMID: 16447092] Acetylcysteine and contrast agent-associated nephrotoxicity. J Am Coll Cardiol.
9. Solomon R, Briguori C, Biguori C, Bettmann M. Selection of contrast me-
dia. Kidney Int Suppl. 2006:S39-45. [PMID: 16612400] 30. Coyle LC, Rodriguez A, Jeschke RE, Simon-Lee A, Abbott KC, Taylor AJ.
10. McCullough PA, Wolyn R, Rocher LL, Levin RN, O'Neill WW. Acute
Acetylcysteine In Diabetes (AID): a randomized study of acetylcysteine for the renal failure after coronary intervention: incidence, risk factors, and relationship prevention of contrast nephropathy in diabetics. Am Heart J. 2006;151:1032.e9- to mortality. Am J Med. 1997;103:368-75. [PMID: 9375704] 11. Barrett BJ, Carlisle EJ. Metaanalysis of the relative nephrotoxicity of high-
31. Diez T, Bagilet D, Ramos M, Jolly H, Diab M, Marcucci R, et al. [Eval-
and low-osmolality iodinated contrast media. Radiology. 1993;188:171-8.
uation of two methods to avoid the nephropathy associated with radiologic con- trast]. Medicina (B Aires). 1999;59:55-8. [PMID: 10349120] 12. Aspelin P, Aubry P, Fransson SG, Strasser R, Willenbrock R, Berg KJ;
32. Diaz-Sandoval LJ, Kosowsky BD, Losordo DW. Acetylcysteine to prevent
Nephrotoxicity in High-Risk Patients Study of Iso-Osmolar and Low-Osmolar
angiography-related renal tissue injury (the APART trial). Am J Cardiol. 2002; Non-Ionic Contrast Media Study Investigators. Nephrotoxic effects in high-risk
patients undergoing angiography. N Engl J Med. 2003;348:491-9. [PMID: 33. Drager LF, Andrade L, Barros de Toledo JF, Laurindo FR, Machado Ce'sar
LA, Seguro AC. Renal effects of N-acetylcysteine in patients at risk for contrast
13. Eisenberg RL, Bank WO, Hedgock MW. Renal failure after major angiog-
nephropathy: decrease in oxidant stress-mediated renal tubular injury. Nephrol raphy can be avoided with hydration. AJR Am J Roentgenol. 1981;136:859-61.
Dial Transplant. 2004;19:1803-7. [PMID: 15128882] 34. Durham JD, Caputo C, Dokko J, Zaharakis T, Pahlavan M, Keltz J, et al.
14. Jo SH, Youn TJ, Koo BK, Park JS, Kang HJ, Cho YS, et al. Renal toxicity
A randomized controlled trial of N-acetylcysteine to prevent contrast nephropa- evaluation and comparison between visipaque (iodixanol) and hexabrix (ioxag- thy in cardiac angiography. Kidney Int. 2002;62:2202-7. [PMID: 12427146] late) in patients with renal insufficiency undergoing coronary angiography: the 35. Dussol B, Morange S, Loundoun A, Auquier P, Berland Y. A randomized
RECOVER study: a randomized controlled trial. J Am Coll Cardiol. 2006;48: trial of saline hydration to prevent contrast nephropathy in chronic renal failure patients. Nephrol Dial Transplant. 2006;21:2120-6. [PMID: 16611682] 15. McCullough PA, Bertrand ME, Brinker JA, Stacul F. A meta-analysis of the
36. El Mahmoud R, Le Feuvre C, Le Quan Sang KH, Helft G, Beygui F,
renal safety of isosmolar iodixanol compared with low-osmolar contrast media. J Batisse JP, et al. [Absence of nephro-protective effect of acetylcysteine in patients
Am Coll Cardiol. 2006;48:692-9. [PMID: 16904536] with chronic renal failure investigated by coronary angiography]. Arch Mal Coeur 16. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improv-
ing the quality of reports of meta-analyses of randomised controlled trials: the 37. Erley CM, Duda SH, Rehfuss D, Scholtes B, Bock J, Mu¨ller C, et al.
QUOROM statement. Quality of Reporting of Meta-analyses. Lancet. 1999; Prevention of radiocontrast-media-induced nephropathy in patients with pre- existing renal insufficiency by hydration in combination with the adenosine an- 17. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials.
tagonist theophylline. Nephrol Dial Transplant. 1999;14:1146-9. [PMID: 18. Marenzi G, Assanelli E, Marana I, Lauri G, Campodonico J, Grazi M, et al.
38. Fung JW, Szeto CC, Chan WW, Kum LC, Chan AK, Wong JT, et al.
N-acetylcysteine and contrast-induced nephropathy in primary angioplasty. N Effect of N-acetylcysteine for prevention of contrast nephropathy in patients with Engl J Med. 2006;354:2773-82. [PMID: 16807414] moderate to severe renal insufficiency: a randomized trial. Am J Kidney Dis.
19. Spargias K, Adreanides E, Giamouzis G, Karagiannis S, Gouziouta A,
Manginas A, et al. Iloprost for prevention of contrast-mediated nephropathy in
39. Goldenberg I, Shechter M, Matetzky S, Jonas M, Adam M, Pres H, et al.
high-risk patients undergoing a coronary procedure. Results of a randomized pilot Oral acetylcysteine as an adjunct to saline hydration for the prevention of con- study. Eur J Clin Pharmacol. 2006;62:589-95. [PMID: 16832680] trast-induced nephropathy following coronary angiography. A randomized con- 20. Stevens MA, McCullough PA, Tobin KJ, Speck JP, Westveer DC, Guido-
trolled trial and review of the current literature. Eur Heart J. 2004;25:212-8.
Allen DA, et al. A prospective randomized trial of prevention measures in pa-
tients at high risk for contrast nephropathy: results of the P.R.I.N.C.E. Study.
40. Gomes VO, Poli de Figueredo CE, Caramori P, Lasevitch R, Bodanese LC,
Prevention of Radiocontrast Induced Nephropathy Clinical Evaluation. J Am Arau'jo A, et al. N-acetylcysteine does not prevent contrast induced nephropathy
Coll Cardiol. 1999;33:403-11. [PMID: 9973020] after cardiac catheterisation with an ionic low osmolality contrast medium: a 21. Verhagen AP, de Vet HC, de Bie RA, Kessels AG, Boers M, Bouter LM,
multicentre clinical trial. Heart. 2005;91:774-8. [PMID: 15894775] et al. The Delphi list: a criteria list for quality assessment of randomized clinical
41. Gulel O, Keles T, Eraslan H, Aydogdu S, Diker E, Ulusoy V. Prophylactic
trials for conducting systematic reviews developed by Delphi consensus. J Clin acetylcysteine usage for prevention of contrast nephropathy after coronary angiog- Epidemiol. 1998;51:1235-41. [PMID: 10086815] raphy. J Cardiovasc Pharmacol. 2005;46:464-7. [PMID: 16160598] 22. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency
42. Huber W, Ilgmann K, Page M, Hennig M, Schweigart U, Jeschke B, et al.
in meta-analyses. BMJ. 2003;327:557-60. [PMID: 12958120] Effect of theophylline on contrast material-nephropathy in patients with chronic 23. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis
renal insufficiency: controlled, randomized, double-blinded study. Radiology.
detected by a simple, graphical test. BMJ. 1997;315:629-34. [PMID: 9310563] 24. Abizaid AS, Clark CE, Mintz GS, Dosa S, Popma JJ, Pichard AD, et al.
43. Huber W, Schipek C, Ilgmann K, Page M, Hennig M, Wacker A, et al.
Effects of dopamine and aminophylline on contrast-induced acute renal failure Effectiveness of theophylline prophylaxis of renal impairment after coronary an- www.annals.org
19 February 2008 Annals of Internal Medicine Volume 148 • Number 4 293
Review Effectiveness of Drugs for Preventing Contrast-Induced Nephropathy giography in patients with chronic renal insufficiency. Am J Cardiol. 2003;91: 56. Sinha SK, Berry WA, Bueti J, Junaid A, Fine A, Krahn J. The prevention of
radiocontrast-induced nephropathy trial (print): a prospective, double-blind, ran- 44. Jo SH, Koo BK, Youn TJ, Hahn JY, Kim YS, Kang HJ. Prevention of
domized, controlled trial of iso-osmolar versus low-osmolar radiocontrast in com- contrast induced nephropathy by short term statin in patients with renal insuffi- bination with N-acetylcysteine versus placebo [Abstract]. Presented at the Amer- ciency undergoing coronary angiography: a randomized controlled trial [Ab- ican Heart Association Scientific Sessions, New Orleans, Louisiana, 7-10 stract]. In: TCT 2005. Abstracts of the 17th Annual Transcatheter Cardiovascu- November 2004. Accessed at www.abstractsonline.com/arch/RecordPrintView lar Therapeutics Symposium, October 17-21, 2005, Washington DC, USA. Am .aspx?LookupKeyϭ12345&RecordIDϭ717 on 19 December 2007.
J Cardiol. 2005;96:1H-213H. [PMID: 16252422] 57. Stone GW, McCullough PA, Tumlin JA, Lepor NE, Madyoon H, Murray
45. Kapoor A, Kumar S, Gulati S, Gambhir S, Sethi RS, Sinha N. The role of
P, et al; CONTRAST Investigators. Fenoldopam mesylate for the prevention of
theophylline in contrast-induced nephropathy: a case-control study. Nephrol contrast-induced nephropathy: a randomized controlled trial. JAMA. 2003;290: Dial Transplant. 2002;17:1936-41. [PMID: 12401850] 46. Kay J, Chow WH, Chan TM, Lo SK, Kwok OH, Yip A, et al. Acetylcys-
58. Tepel M, van der Giet M, Schwarzfeld C, Laufer U, Liermann D, Zidek
teine for prevention of acute deterioration of renal function following elective W. Prevention of radiographic-contrast-agent-induced reductions in renal func-
coronary angiography and intervention: a randomized controlled trial. JAMA.
tion by acetylcysteine. N Engl J Med. 2000;343:180-4. [PMID: 10900277] 59. Spargias K, Alexopoulos E, Kyrzopoulos S, Iokovis P, Iacovis P, Green-
47. Kefer JM, Hanet CE, Boitte S, Wilmotte L, De Kock M. Acetylcysteine,
wood DC, et al. Ascorbic acid prevents contrast-mediated nephropathy in pa-
coronary procedure and prevention of contrast-induced worsening of renal func- tients with renal dysfunction undergoing coronary angiography or intervention.
tion: which benefit for which patient? Acta Cardiol. 2003;58:555-60. [PMID: Circulation. 2004;110:2837-42. [PMID: 15492300] 60. Webb JG, Pate GE, Humphries KH, Buller CE, Shalansky S, Al Shamari
48. MacNeill BD, Harding SA, Bazari H, Patton KK, Colon-Hernadez P,
A, et al. A randomized controlled trial of intravenous N-acetylcysteine for the
DeJoseph D, et al. Prophylaxis of contrast-induced nephropathy in patients un-
prevention of contrast-induced nephropathy after cardiac catheterization: lack of dergoing coronary angiography. Catheter Cardiovasc Interv. 2003;60:458-61.
effect. Am Heart J. 2004;148:422-9. [PMID: 15389228] 61. Alonso A, Lau J, Jaber BL, Weintraub A, Sarnak MJ. Prevention of radio-
49. Merten GJ, Burgess WP, Gray LV, Holleman JH, Roush TS, Kowalchuk
contrast nephropathy with N-acetylcysteine in patients with chronic kidney dis- GJ, et al. Prevention of contrast-induced nephropathy with sodium bicarbonate:
ease: a meta-analysis of randomized, controlled trials. Am J Kidney Dis. 2004;43: a randomized controlled trial. JAMA. 2004;291:2328-34. [PMID: 15150204] 50. Namgung J, Doh JH, Lee SY, Lee WR. Effect of N acetylcysteine in the
62. Birck R, Krzossok S, Markowetz F, Schnu¨lle P, van der Woude FJ, Braun
prevention of contrast induced nephropathy after coronary angiography [Ab- C. Acetylcysteine for prevention of contrast nephropathy: meta-analysis. Lancet.
stract]. In: Abstracts of the 10th Annual Interventional Vascular TherapeuticsAngioplasty Summit-Transcatheter Cardiovascular Therapeutics Asia Pacific Symposium. April 28 -30, 2005, Seoul, Korea. Am J Cardiol. 2005;95:1A-83A.
63. Liu R, Nair D, Ix J, Moore DH, Bent S. N-acetylcysteine for the prevention
of contrast-induced nephropathy. A systematic review and meta-analysis. J Gen 51. Ochoa A, Pellizzon G, Addala S, Grines C, Isayenko Y, Boura J, et al.
Intern Med. 2005;20:193-200. [PMID: 15836554] Abbreviated dosing of N-acetylcysteine prevents contrast-induced nephropathy 64. Ix JH, McCulloch CE, Chertow GM. Theophylline for the prevention of
after elective and urgent coronary angiography and intervention. J Interv Cardiol.
radiocontrast nephropathy: a meta-analysis. Nephrol Dial Transplant. 2004;19: 52. Oldemeyer JB, Biddle WP, Wurdeman RL, Mooss AN, Cichowski E,
65. Bagshaw SM, Ghali WA. Theophylline for prevention of contrast-induced
Hilleman DE. Acetylcysteine in the prevention of contrast-induced nephropathy
nephropathy: a systematic review and meta-analysis. Arch Intern Med. 2005;165: after coronary angiography. Am Heart J. 2003;146:E23. [PMID: 14661012] 53. Rashid ST, Salman M, Myint F, Baker DM, Agarwal S, Sweny P, et al.
66. Trivedi HS, Moore H, Nasr S, Aggarwal K, Agrawal A, Goel P, et al. A
Prevention of contrast-induced nephropathy in vascular patients undergoing an- randomized prospective trial to assess the role of saline hydration on the devel- giography: a randomized controlled trial of intravenous N-acetylcysteine. J Vasc opment of contrast nephrotoxicity. Nephron Clin Pract. 2003;93:C29-34.
54. Sandhu C, Belli AM, Oliveira DB. The role of N-acetylcysteine in the
67. McAuley L, Pham B, Tugwell P, Moher D. Does the inclusion of grey
prevention of contrast-induced nephrotoxicity. Cardiovasc Intervent Radiol.
literature influence estimates of intervention effectiveness reported in meta-anal- yses? Lancet. 2000;356:1228-31. [PMID: 11072941] 55. Shyu KG, Cheng JJ, Kuan P. Acetylcysteine protects against acute renal
68. Cook DJ, Guyatt GH, Ryan G, Clifton J, Buckingham L, Willan A, et al.
damage in patients with abnormal renal function undergoing a coronary proce- Should unpublished data be included in meta-analyses? Current convictions and dure. J Am Coll Cardiol. 2002;40:1383-8. [PMID: 12392825] controversies. JAMA. 1993;269:2749-53. [PMID: 8492400] 294 19 February 2008 Annals of Internal Medicine Volume 148 • Number 4
www.annals.org
Annals of Internal Medicine
Current Author Addresses: Drs. Kelly and Cronin: Department of
Dr. Bernstein: Department of Internal Medicine, University of Michi- Radiology, Division of Cardiothoracic Imaging, University of Michigan, gan, 1500 East Medical Center, Ann Arbor, MI 48109.
1500 East Medical Center, Ann Arbor, MI 48109.
Dr. Carlos: Department of Radiology, Division of Magnetic Resonance Dr. Dwamena: Division of Nuclear Medicine, University of Michigan, Imaging, University of Michigan, 1500 East Medical Center, Ann Arbor, 1500 East Medical Center, Ann Arbor, MI 48109.
Appendix Table. Summary of Study Quality Characteristics
Study, Year (Reference)
Concealed
Specific
Estimate
Intention-to-
Allocation
Baseline
Inclusion
Variability of
Criteria
Assessor
Provider
Outcomes
Analysis
W-58 19 February 2008 Annals of Internal Medicine Volume 148 • Number 4
www.annals.org

Source: http://www.emottawa.ca/assets_secure/journal_club/articles/Preventing%20Contrast%20Induced%20Nephropathy_Kelly_AnnalsIM_080219%20(2).pdf