Objective Previous studies have suggested that statin pretreatment prevents contrast-induced nephropathy (CIN). However, single randomised trials are limited in their number of patients. This meta-analysis aims to assess the role of statin use in CIN prevention, as well as to determine patient subgroups that will benefit from statin pre-treatment.
Methodology We searched PubMed, EMBASE, Web of science and the Cochrane Central Register of Controlled Trials databases for randomised controlled trials (RCT) comparing statin pretreatment versus placebo for preventing CIN. Our main outcome was the risk of CIN within 1–5 days after contrast administration.
Results Data analysed from nine randomised studies with a total of 5143 patients, where 2559 received statins and 2584 received placebo, showed that statin pretreatment was associated with significant reduction in risk of CIN (MH-RR=0.47, 95% CI 0.34 to 0.64, Z=4.49, p<0.00001). This beneficial effect of statin was also seen in patients with baseline renal impairment (MH-RR=0.46, 95% CI 0.29 to 0.72, p=0.0008) and also those who were cotreated with NAC (MH-RR=0.46, 95% CI 0.25 to 0.83, p=0.01).
Conclusions Statin pretreatment leads to significant reduction in CIN, and should be strongly considered in all patients who are planned for diagnostic and interventional procedures involving contrast-media administration.
Statistics from Altmetric.com
What is already known about this subject?
Previous studies have suggested that statin pre-treatment prevents contrast-induced nephropathy (CIN).
However, single randomized trials are limited in their number of patients, and there is still a lack of conclusive evidence on statin pre-treatment to prevent CIN.
What does this study add?
This meta-analysis of nine randomized trials demonstrated that statin pre-treatment leads to significant reduction in CIN in patients undergoing coronary angiography, and is independent of renal impairment, type of contrast used, and concomitant N-acetyl cysteine therapy.
How might this impact on clinical practice?
Our findings suggest strong consideration of statin pre-treatment in all patients who are planned for diagnostic and interventional procedures involving contrast-media administration.
This meta-analysis sets a platform for future pharmacological guidelines and clinical trials, as well as highlights the need for trials studying statin’s efficacy in other non-cardiac contrast procedures such as radiological imaging.
Contrast-induced nephropathy (CIN) or contrast-induced acute kidney injury (CI-AKI) is the third most common cause of acute renal injury in the hospital setting.1 ,2 CIN is defined as an increase of serum creatinine of 0.5 mg/dL or a 25% relative increase in creatinine at 48 h after contrast exposure.3 CIN has been associated with poor short-term and long-term outcomes.4 Patients who develop CIN are predisposed to a greater risk of developing chronic kidney disease (CKD) in the long term.5 This condition is also associated with increased morbidity, lengthened hospital stays and increased healthcare costs.6 Therefore, increasing efforts have been made to identify effective preventive strategies for CIN. The best approach to prevent CIN remains unclear. Current recommendations only include adequate preparatory hydration.7 There is still a lack of conclusive evidence in other strategies such as administration of N-acetylcysteine (NAC) as an antioxidant, sodium bicarbonate-based hydration and consideration of the type of contrast used.3
The pathophysiology of CIN involves renal vasoconstriction, inflammation and oxidative stress.8 These mechanisms have led to the proposed use of hydroxymethylglutaryl coenzyme A reductase inhibitors, otherwise known as statins, for the prevention of CIN.
There have been multiple randomised controlled trials (RCTs) performed to evaluate the efficacy of statin pretreatment in the prevention of CIN, with earlier trials failing to strongly demonstrate the beneficial effects of statin. Multiple meta-analyses have examined the role of statin pretreatment, but they had limited, variable and conflicting conclusions given the limited number of RCTs performed.9–12 Recently, the results of two large RCTs were published,13 ,14 adding more data into the existing pool and possibly providing new insights into the efficacy of statin to prevent contrast-nephropathy.
We therefore conducted a meta-analysis of RCTs to further evaluate the efficacy of statin pretreatment for prevention of CIN, as well as to ascertain its effectiveness in various patient subgroups.
A systematic literature review was planned and performed using methods specified in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for systematic review. Both controlled vocabulary terms (eg, MeSH) and key words were used to search the following databases for articles related to the role of statins in preventing or ameliorating CIN: MEDLINE/PubMed, EMBASE, Cochrane Library, Web of Science and ClinicalTrials.gov. Literature searches were completed in March, 2014. The complete MEDLINE/PubMed Search strategy, on which the other database searches were also built, is available in online supplementary appendix A. Reference lists of citations to the ultimately included articles were also searched for articles that would meet inclusion criteria, as was the grey literature, including the websites of national and international nephrology societies.
Study selection, data extraction and quality assessment
Our prespecified inclusion criteria were as follows: (1) randomised controlled trials, (2) treatment arm containing any of the available statins alone or with NAC, (3) control arm containing placebo or NAC, (4) both arms administer NAC if the study incorporated NAC, (5) intravenous or intra-arterial iodinated contrast medium. Our primary outcome was the development of contrast-induced nephropathy, defined as an absolute increase of serum creatinine of 0.5 mg/dL or an increase in baseline serum creatinine level of 25% within 24 h to 5 days after exposure to contrast medium or serum Cystatin concentration >10% within 24 h after exposure to contrast medium.15 Secondary outcome measurement was the need for dialysis within 24 h to 5 days after contrast medium exposure.
Two independent reviewers performed the study selection (NS, JL). In case of disagreements, a third reviewer (CH) cast the deciding vote. Titles and abstracts of retrieved references were screened for inclusion and full texts of potential articles were further analysed to see if they met inclusion criteria. Case reports, letters, systematic reviews/meta-analyses and data published in the form of abstracts without peer-reviewed publication of the manuscripts were excluded. Articles not in English were not automatically excluded though four non-English articles were ultimately excluded due to the absence of adequate translation. Both individuals who collected the data used the following study specific characteristics, including study name, sample size of treatment group and control group per intention to treat analysis, event number and rate in both the treatment group and control group, statistical effect estimates used in the individual studies, type of statin used, statin protocol used, hydration protocol, NAC use, type and amount of contrast medium used, procedures which patients underwent, inclusion and exclusion criteria of individual studies, CIN definition prespecified in individual studies, rate of adverse events in treatment group and control group, and baseline patient characteristics each study including presence of diabetes, history of statin use and degree of renal impairment. Quality assessment was judged on allocation concealment, randomisation, blinding, completion of follow-up and intention-to-treat analysis.16 Study quality was assessed following the pattern of PRISMA guidelines and explanation.17 Attempts were made to contact the corresponding authors of included studies for any additional information, which was deemed relevant to this meta-analysis.
Statistical analyses were performed using Review Manager V.5.2.9 software (available from The Cochrane Collaboration). A study or trial level pooled analysis of the included RCT’s was performed to evaluate the effect of pretreatment with statin on prevention of CIN. The analysis was performed according to the intention-to-treat strategy. Mantel Haenszel risk ratio (MH RR) with 95% CI was calculated. The χ2 statistic was calculated and a formal test of heterogeneity was conducted. The I2 index was used to summarise the proportion of the total variability in the estimates due to between-study variation. We regarded I2 of less than 25%, 25–50% and >50% as low, moderate, and high amounts of heterogeneity, respectively.18 Random-effects model rather than the fixed-effects model was used to assess the overall estimate.19 We assessed for potential publication bias by using the funnel plots of SE of RR versus RR.20 Subgroup analysis was conducted based on renal impairment of participants, type of contrast medium (iodixanol (iso-osmolar contrast) or non-iodixanol (low-osmolar contrast)), history of diabetes mellitus (with or without diabetes mellitus), type of statin used (atorvastatin or rosuvastatin), and the addition of NAC (with or without NAC). We also analysed the relative risk of acute renal failure requiring dialysis. All the tests were two tailed and a p value less than 0.05 was regarded as significant in this meta-analysis.
A total of 2173 potentially relevant citations were identified. Figure 1 shows our search strategy, which yielded nine randomised controlled trials with 5143 patients.13 ,14 ,21–26 A total 2559 patients were randomised to receive statins, and 2584 patients to receive the placebo or NAC alone.
Patient characteristics and interventions
Among the nine RCTs, Jo et al26 used 40 mg of simvastatin as a treatment arm against placebo while six other studies21–25 ,27 used atorvastatin; 80 mg atorvastatin was used in five studies22–25 and 40 mg was used in one study.21 Two recent trials used rosuvastatin in different doses, with Han et al13 using 10 mg and Leoncini et al14 using 40 mg. Both studies showed significant beneficial effect of statin pretreatment. In terms of contrast medium, four studies21 ,24 ,25 ,27 used non-ionic low-osmolar contrast medium and the rest of the studies13 ,14 ,23 ,26 used non-ionic iso-osmolar contrast medium.
The contrast procedure performed were percutaneous coronary intervention (PCI) in two studies,24 ,27 coronary angiography or percutaneous coronary in three studies,14 ,22 ,23 coronary angiography only in three studies,21 ,25 ,26 and coronary angiography, percutaneous coronary intervention or peripheral arterial angiography in one study.13
The majority of the studies used serum creatinine increase of >0.5 mg/dL within 24 h to 5 days as their definition of CIN with exception of two studies.21 ,23 Quintavalle et al23 used serum Cystatin C increase (10%) within 24 h of contrast medium administration and Acikel et al21 used combination of serum creatinine, Cystatin C and glomerular filtration rate as the definition for CIN.
One similarity among all the studies was the inclusion of patients with diabetes in both arms of the RCTs. Of the nine studies, six studies13 ,14 ,22–24 ,26 had patients with chronic kidney disease with creatinine clearance <60 mL/min or glomerular filtration rate (GFR) <60 mL/min/1.73 m2 and two studies21 ,25 excluded patients with GFR <70 mL/min/1.73 m2. Li et al27 also excluded patients with history of renal dysfunction.
The primary end point of eight out of the nine studies was CIN, defined as an increase in serum creatinine by more than 25% or 0.5 mg/dL within 3 days, or serum Cystatin concentration >10% within 24 h of the administration of contrast medium in the absence of alternative explanations.28 In the Jo et al26 study, change in serum creatinine level was the primary end point and incidence of CIN was the secondary end point.
Based on pooled estimates across the nine randomised controlled trials, there was a significant benefit associated with statin treatment (MH-RR=0.47, 95% CI 0.34 to 0.64, p<0.00001). Thus, statin use, irrespective of type of statin, has a beneficial role in preventing CIN. We drew our conclusions using the random-effects model for this analysis given the clinical and methodological heterogeneity of the RCTs included in the meta-analysis.19
The Forest plot in figure 2 shows the results of studies pooled in the meta-analysis.
There was moderate heterogeneity among the studies using random method (I2=29%, χ2=11.24, df=8, p=0.19). The I2 measure was used to quantify heterogeneity as it is independent of the number of studies in the meta-analysis and has better power as compared to using the χ2 method.29 ,30
Sensitivity analysis measured the effect of individual studies on the summary effect size. It showed that the summary effect was statistically significant (MH-RR=0.40, 95% CI 0.28 to 0.59, p<0.00001, benefit by 7%). After removing studies by Ozhan et al,25 Toso et al,22 Jo et al26 study and Acikel et al,21 the statistical significance improved. Publication bias was assessed by Funnel plot (figure 3) with SE of log RR against RR. It shows the symmetrical distribution of the plot around the summary effect size showing no publication bias.
There is clinical and methodological diversity in the studies included in this meta-analysis. This issue was addressed by using subgroup analysis as prespecified in the methods section.
Out of nine clinical trials, six studies13 ,14 ,22–24 ,26 included patients with pre-existing renal impairment. Among them, two14 ,24 have patients with normal renal function and the other five13 ,22 ,23 ,26 (8, 9, 12, 16) included only patients without renal impairment at the baseline. Ozhan et al21 ,25 excluded patients with GFR <70 min/mL/1.73 m2. The definition of pre-existing renal impairment among the trials was not consistent. Four studies14 ,22 ,24 ,26 included patients with serum creatinine clearance <60 mL/min; and the remaining trials13 ,23 only included GFR >60 mL/min/1.73 m2. Therefore, our study defined renal impairment as GFR <60 mL/min/1.73 m2 or creatinine clearance <60 mL/min, excluding patients on dialysis.
In our subgroup analysis, we found that statins significantly reduced the risk of CIN in patientswith or without renal impairment (figure 4). In patients with renal impairment, statin reduced CIN risk by 54% (MH-RR=0.46, 95% CI 0.29 to 0.72, p=0.0008). In patients without renal impairment, statin reduced the CIN risk by 55% (MH-RR=0.42, 95% CI 0.28 to 0.63, p<0.0001). There was no statistically significant difference in the degree of beneficial effect of statin on CIN prevention between both groups (p=0.79).
Acute renal failure requiring dialysis in statin versus control group
Out of nine RCTs, four trials13 ,14 ,22 ,26 defined the number of patients who developed acute renal failure requiring dialysis. In Toso et al22 and Jo et al26 trials, the statin groups had no patients developing acute renal failure requiring dialysis, whereas the control groups had one event each. In two trials,14 ,13 there were two events of acute renal failure requiring dialysis in the control group compared to none in the statin group. We analysed the relative risk of this major adverse event in the studies mentioned above, but it was not statistically significant (MH-RR=0.25, 95% CI 0.05 to 1.21, p=0.09).
Quintavalle et al23 described the need for chronic dialysis in eight patients without differentiating its occurrence in either statin group or control group, from a total of 37 major events in 402 patients.
Low-osmolar versus iso-osmolar contrast media
The CIN prevention effect of statin therapy was also found to be independent of the type of contrast medium used (figure 5). Statin use significantly decreased the incidence of CIN by 48% in groups of patient who received iso-osmolar contrast (MH-RR=0.52, 95% CI 0.34 to 0.79, p=0.002) and 64% in low-osmolar contrast (MH-RR=0.31, 95% CI 0.16 to 0.60, p=0.0006). There was no statistically significant difference between the two groups in the degree of statin's CIN prevention effect (p=0.19).
Concomitant pretreatment with NAC
Among the nine RTC trials, four of the trials used NAC as a part of the study protocol in both treatment arm and control arm. Statin's CIN prevention effect was found to be independent of NAC use (figure 6). In patients without concurrent NAC therapy, statin reduced the risk of CIN by 49% (MH-RR=0.51, 95% CI 0.36 to 0.73, p=0.0002). In patients who received NAC, statin reduced the risk of CIN by 54% (MH-RR=0.46, 95% CI 0.25 to 0.83, p=0.01). The efficacy of statins in prevention of CIN between both groups was not statistically different (p=0.74).
Statin type (comparison between atorvastatin and rosuvastatin)
From the nine RCT, there were six studies that used atorvastatin and two newer trials Han et al13 and Leoncini et al14 used rosuvastatin. The beneficial effect of statin was seen in patients who received atorvastatin (MH-RR=0.38, 95% CI 0.21 to 0.71, p=0.002) as well as patients who received rosuvastatin (MH-RR=0.53, 95% CI 0.38 to 0.74, p=0.0002; figure 7).
In our subgroup analysis, we compared the outcome of interest (CIN) based on diabetes mellitus status (figure 8). The risk reduction of CIN in patients with diabetes was 39% (MH-RR=0.61, 95% CI 0.43 to 0.86, p=0.005), and in patients without diabetes was 39% (MH=RR=0.61, 95% CI 0.33 to 1.13, p=0.11). However, the difference in risk reduction of CIN was not statistically significant between both groups (p=0.99).
This meta-analysis based on nine randomised controlled trials demonstrates that administration of statin therapy has a statistically significant benefit of preventing CIN among patients undergoing interventional procedures requiring contrast media when compared to placebo or NAC. In addition, both rosuvastatin and atorvastatin has shown statistically beneficial role in preventing CIN.
CIN is a recognised complication after contrast media administration during interventional procedures. The beneficial effect of statins on CIN prevention may be explained by multiple mechanisms. After exposure to contrast medium, there is a short period of vasodilation followed by intense renal vasoconstriction. Various molecules, including angiotensin, appear to be involved in the mechanism of this renal hypoperfusion.31 Statins down-regulate angiotensin receptors and decrease endothelin synthesis, and therefore may prevent renal hypoperfusion and ischaemia.32 Contrast medium may also cause direct damage to tubular cells facilitated by oxygen-free radicals, proinflammatory cytokines and complement activation, leading to protein precipitates, which causes tubular obstruction.31 Statins have antioxidant and anti-inflammatory properties as well as tendency to reduce endothelin secretion, which may be the mechanism for the CIN prevention effect of statins.33 ,34
Our present study adds to the current literature by suggesting that statins prevent CIN even in patients with pre-existing renal impairment. Previous studies and meta-analysis were not able to find any CIN preventive effect of statins in patients with chronic kidney disease.9 ,22 ,26 ,35 This is most likely because of the smaller sample size of patients being studied and analysed. We included a total of 1330 patients with renal impairment in our subgroup analysis of the CIN prevention effect of statins. This is concordant with other studies showing the beneficial effects of statins in patients with CKD.36 ,37 However, given that most of our included trials have excluded patients with GFR <30, it remains unclear if statins will be beneficial in reducing CI-AKI in patients with CKD stage 4 or 5. We also found that there were more patients who developed CIN, who eventually required dialysis in the control group as compared to the group who received statin therapy (six out of 2022 vs none out of 2019), although conclusions from this observation cannot be made due to incomplete reporting of ultimate progression to haemodialysis from several of the studies.
Our analysis also showed that the CIN preventive effect of statins is seen in patients receiving iso-osmolar or low-osmolar contrast. This finding is in keeping with the 2012 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines,2 which recommend both options of contrast (low-osmolal or iso-osmolar) without a precedence of one over the other.38 Looking at the absolute reduction of CIN in our analysis, low-osmolar contrast had a lower incidence of CIN. However, this absolute reduction of CIN might be attributed to the reduced average amount of contrast used when compared to iso-osmolar contrast.
Risk factors for CIN include CKD, use of first generation hyperosmolal ionic contrast agents, hypovolemia, non-steroidal anti-inflammatory drug (NSAID) use, renal transplantation and heart failure.39–41
In our subgroup analysis, we found the CIN prevention effect of statins was independent of the use of NAC. It has been previously suggested that statins would not benefit patients who had been premedicated with NAC, with the hypothesis that both statins and NAC act on similar pathways.9 Our findings suggest that patients may benefit from administration of statins regardless of whether they were planned to receive NAC or not.
We also found that the beneficial effects of statins were seen equally in patients that were treated with high-dose atorvastatin and rosuvastatin. Definitive conclusions regarding the superiority of one of the two mentioned statins over the other cannot be made. The question of whether simvastatin exerts the same CIN prevention effect remains unclear, especially when the RCTs using simvastatins are limited. In Jo et al,26 simvastatin did not show a statistical significant decrease risk of CIN. This raises the question on the role of statin potency in the mechanism of CIN prevention.
Our study yielded a significant result in statin's CIN prevention efficacy in patients with diabetes. Our non-significant result in patients with no diabetes is most likely due to inadequate sample size in the group with no diabetes, as our subgroup analysis included 3228 patients with diabetes and only 914 patients without diabetes. Supporting this theory is that both groups were found to have similar RRs of 0.61.
There are several limitations in this meta-analysis. First, during the analysis, we found that most of the randomised controlled trials of statin use as prevention of CIN were performed in patients undergoing coronary angiography or PCI. We are unaware of randomised trials examining statin use for the prevention of CIN outside of cardiovascular angiography. Given likely pre-existing use in this patient cohort, the finding of statin benefit is unlikely to have large clinical implication. There clearly exists a need to repeat these trials in patients who are likely to be statin naïve, undergoing iodinated contrast exposure for non-cardiovascular imaging procedures such as CT scan.
Second, the studies analysed the CIN prevention effect of statins with a total of three types of statins, at varied doses, for different duration of treatment. The latest two studies used rosuvastatin for 12 h prior to the procedure, and were continued for 12 and 24 h in each study, respectively. These variations in treatment protocols may have lead to different effects in prevention of CIN. Third, the definition of CIN was not the same in all the studies and varied in the amount of absolute increase of creatinine and under different predetermined post-procedural time period.
The results of this meta-analysis suggest that statin pretreatment reduces the incidence of CIN, and is independent presence of renal impairment, of the type of contrast used and concomitant NAC therapy. Further studies are needed to determine the appropriate statin pretreatment regimen, including the type, dosing and duration of statin pretreatment. Our study sets a platform for future pharmacological guidelines and clinical trials, as well as highlights the need for trials studying statin's efficacy in other non-cardiac contrast procedures such as radiological imaging.
- Received March 25, 2014.
- Revision received June 3, 2014.
- Accepted July 15, 2014.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Contributors NS, JZL and JJH were involved in integration of data and drafting of manuscript. SWL took part in revision of the manuscript and integration of data. CH was involved in systematic search of the literature. AP took part in integration of data and revision of manuscript. PS took part in revising the manuscript critically for important intellectual content. KSL was involved in conception and design, revising critically for important intellectual content and final approval of the manuscript.
Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data are available.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.