NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Eng J, Subramaniam RM, Wilson RF, et al. Contrast-Induced Nephropathy: Comparative Effects of Different Contrast Media [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2015 Dec. (Comparative Effectiveness Reviews, No. 155.)
Contrast-Induced Nephropathy: Comparative Effects of Different Contrast Media [Internet].
Show detailsIn this systematic review of the comparative effects of different types of contrast media with respect to developing CIN, we found two types of RCTs: trials comparing two or more LOCMs to each other, and trials comparing IOCM to a LOCM. The small number of trials comparing LOCMs reported no statistically significant or clinically important differences for heterogeneously defined endpoints for CIN. The strength of evidence for the comparison of LOCMs was low, primarily due to the small number of available studies. For the trials comparing IOCM to LOCMs, we found a slight reduction in CIN risk for IOCM that was of borderline statistical significance, with a 95% CI: 0.65 to 0.99 for the relative risk. However, the point estimate of the pooled relative risk reduction (0.80) did not exceed a minimally important relative risk difference of 25 percent. The strength of evidence for the comparison of IOCM to LOCMs was moderate rather than high because most studies of this comparison had either medium or high study limitations despite exceeding the optimum information size.
The majority of trials in our review involved patients receiving intra-arterial administration of contrast. In the small number of trials involving intravenous administration, we saw no evidence that the relationship between contrast type and CIN risk differed from that observed in the intra-arterial trials. It should be noted that this finding represents the lack of an association between route of administration and the comparative risk between contrast types. This is not the same as the simpler relationship between route of administration and absolute CIN risk, which was not encompassed by our systematic review. Narrative reviews of the CIN literature have suggested that intravenous administration is safer than intra-arterial,59 but we did not find evidence that the comparative CIN risk of different types of contrast media varies by route of administration.
In our systematic review, we sought evidence on the relationship between contrast type and renal function. Therefore, our inclusion criteria focused on CIN as the primary outcome under consideration. We collected data on other outcomes of interest, however. Since the majority of studies involved coronary artery procedures, cardiovascular event outcomes were of particular interest. A recent meta-analysis of RCTs compared IOCM and LOCM with cardiovascular events as a reported outcome,60 and found no conclusive evidence that IOCM is superior to LOCM with respect to cardiovascular events. Our review likewise found no conclusive evidence for a difference with respect to cardiovascular events, mortality, subsequent need for renal replacement therapy, or other adverse events. It is important to note, however, that our review of the differences between types of contrast media was part of a comprehensive review that focused primarily on assessing the comparative effectiveness of interventions for preventing CIN.61 Thus, our inclusion criteria targeted trials that were designed to examine the effects of interventions and types of contrast media on the risk of CIN. Therefore, our review may not have included some studies that focused on the effects of different types of contrast media on clinical outcomes other than the risk of CIN. For example, the recent meta-analysis of cardiovascular events by Zhang60 included four RCTs (out of 11) which did not report outcomes directly related to CIN. The evidence grades we assigned to outcomes other than CIN apply only to evidence from studies reporting CIN and do not necessarily apply to all studies reporting these non-renal outcomes.
Our results and summary relative risks are similar to three published meta-analyses which reported no statistically significant reduction of CIN with IOCM compared to LOCM.62-64 Even though our review included six RCTs that have been published since those three meta-analyses, we obtained a similar summary relative risk and 95% CI. This similarity enhances our confidence in concluding that IOCM does confer a small reduction in CIN that is not clinically important.
Although five previously published systematic reviews examining trials comparing IOCM against LOCM have reported statistically significant results favoring IOCM, we identified reasons for the discrepancy from our results. In the case of one of those meta-analyses,65 the two studies favoring IOCM the greatest66,67 were excluded from our analysis because CIN was not adequately defined in the two studies. Two other systematic reviews did not strictly evaluate direct comparisons, but employed analytical methods that allowed indirect comparisons of contrast agents across individual studies.68,69 Those two reviews reported differences specifically between IOCM and the LOCM iohexol, but not with other LOCMs. In our meta-analysis, as shown in Figure 3, the two studies that compared iohexol to IOCM were the two oldest studies and were among the four studies reporting the greatest difference favoring IOCM. One of the reviews involved a broadly defined outcome and included studies with outcomes other than CIN.68 The other review pooled data from observational studies with data from RCTs.69 Two other meta-analyses which reported differences between IOCM and LOCMs70,71 may have been affected by inclusion criteria that were different than those used in our review. One of those meta-analyses included only trials of IOCM that were sponsored by its manufacturer.70 The other meta-analysis71 included a large unpublished positive trial comparing IOCM with iopromide in 1656 patients that comprised 28 percent of the subjects in the review. Data for this trial are only available in a 2010 meeting abstract; to date, the study has not been published.
It should be noted that our review addressed a clinical comparison involving contrast media and did not seek to review evidence concerning the pathophysiology, causal pathway, or epidemiology of CIN. The precise mechanism of CIN is not entirely understood. Some evidence exists from propensity-score matched, retrospective studies questioning the strength of the relationship between contrast administration and CIN.6 This relationship is very important for designing future research, but does not affect the conclusions of this review regarding the comparative impact of contrast media type on observed CIN.6,12,72
Limitations of the Evidence
Limitations of the published evidence should be noted. One of the biggest limitations is that the body of evidence is limited by the relatively small size of the available studies, making it difficult to derive precise estimates of any potential differences. Another limitation of the evidence is that few studies in our review reported on clinical outcomes other than the incidence of CIN. Diagnostic and therapeutic procedures involving iodinated contrast media are generally safe, so it is expected that major adverse events would be rare relative to CIN. Therefore, clinical trials may only have sufficient power to detect large differences in the incidence of major adverse events. While CIN was the primary outcome of interest, we collected data on other associated outcomes, such as cardiovascular events, mortality, adverse events, and image quality. Despite their clinical importance, we found these associated outcomes were inconsistently reported or omitted in the literature that we reviewed.
We found that studies examining CIN generally included patients based on referral for a diagnostic or therapeutic procedure and provided little detail about the distribution of specific clinical indications for the procedures or other details related to the clinical setting such as referral patterns, procedure urgency, severity of renal impairment, and other potential risk factors for CIN. Furthermore, details concerning the procedures themselves were commonly omitted, such as total contrast volume, length of procedure, and contrast injection rates, even though these details are considered important elements of the procedures and are commonly recorded. We found that studies examining the risk of CIN with different types of contrast media generally provided little detail about clinical indications for the diagnostic or therapeutic procedures or other relevant clinical details such as the severity of renal impairment. These are all potential sources of unexplained heterogeneity among the studies in our review. Our inclusion criteria did not select studies based on any of these characteristics, so the results likely apply to a relatively diverse population of patients and procedures.
Limitations of the Review
One limitation of the review is that we generally considered LOCM together as a group even though there were seven chemically different LOCMs in the evidence we reviewed. While direct comparisons of LOCMs are sparse, there is some indirect evidence of heterogeneity involving iohexol. The greatest CIN reduction with IOCM was reported in a study comparing it to iohexol.45 As mentioned previously, two indirect comparisons also concluded that differences existed between iohexol and other LOCMs.68,69
The relatively large number of trials comparing IOCM to a LOCM, in theory, provides indirect information about comparisons between LOCMs. We considered whether a network meta-analysis could be performed to combine this indirect information with the data from direct comparisons. However, the sparse number of direct LOCM comparisons compared to indirect comparisons via IOCM severely limits the reliability of such an analysis.73 For this reason, a network meta-analysis was not performed in our review of the evidence.
Future Research
Since we are unable to draw definitive conclusions on how differences in CIN risk associated with contrast type are modified by demographic characteristics, comorbid conditions, baseline renal function, or use of interventions to prevent CIN, there is a need for additional research in this area. These interactions were either not examined in the reviewed studies, or the factors were inconsistently defined or reported. It makes sense to give highest priority to factors most likely to be associated with a high risk of CIN, such as baseline renal dysfunction or comorbid conditions associated with a high risk of kidney disease.
Additional RCTs comparing IOCM and LOCMs with respect to CIN risk would increase the strength of evidence and precision of pooled effect estimates associated with these comparisons. However, since we found that the CIN risk reduction associated with IOCM is relatively small and unlikely to be clinically important, the necessity for increased precision must be justified prior to conducting additional RCTs.
Conclusions
In summary, RCTs comparing LOCMs with each other are relatively sparse, but none reported a statistically significant or clinically important difference with respect to CIN. This absence of a difference is associated with a low strength of evidence. A larger number of trials compared IOCM to LOCM with respect to CIN. In aggregate, these trials demonstrated moderate strength of evidence for a slight CIN reduction associated with IOCM compared to a diverse group of LOCMs. However, this reduction had only borderline statistical significance and is unlikely to be clinically important. No relationship was found between comparative CIN risk and route of administration.
- Introduction - Contrast-Induced NephropathyIntroduction - Contrast-Induced Nephropathy
- Discussion - Contrast-Induced NephropathyDiscussion - Contrast-Induced Nephropathy
- Treatment for Restless Legs Syndrome: Future Research NeedsTreatment for Restless Legs Syndrome: Future Research Needs
- Prevention of Venous Thromboembolism After InjuryPrevention of Venous Thromboembolism After Injury
- Renal Artery Stenosis Management StrategiesRenal Artery Stenosis Management Strategies
Your browsing activity is empty.
Activity recording is turned off.
See more...