Introduction

The intravascular administration of iodinated contrast media has been a recognized etiology of acute kidney injury (AKI) for decades [1]. Recent studies have questioned the causal association of iodinated contrast administration with acute impairment in kidney function [2,3,4,5,6,7]. Review of specific lines of past scientific inquiry on contrast administration and AKI and critical analysis of recent research questioning this association demonstrates that contrast-associated AKI (CA-AKI) is not a myth.

Discussion

The effects of contrast on the kidney

Past research on the physiologic effects of intravascular contrast in the kidney supports the nephrotoxicity of iodinated contrast (Fig. 1). In some animal models, the intravascular administration of iodinated contrast results in decreased renal blood flow to and a reduction in the partial pressure of oxygen of the outer renal medulla, a segment of the kidney that is particularly vulnerable to perturbations in oxygen supply [8, 9]. This adverse hemodynamic effect of contrast was also observed in studies of healthy human subjects using blood oxygen level-dependent MRI, in which the intravascular administration of iodinated contrast reduced renal medullary blood flow. Contrast administration in animals has also been shown to increase the generation of oxygen free radicals, an effect that is associated with a decrease in glomerular filtration [8]. Finally, in vitro studies demonstrate that iodinated contrast has adverse effects on mitochondrial enzyme activity and membrane function and contributes to apoptosis of renal tubular epithelial cells [10, 11]. Although findings in animal models do not necessarily translate to humans and the aforementioned studies in animals have certain methodologic limitations, these and other studies provide a pathophysiologic basis for the nephrotoxicity of iodinated contrast.

Fig. 1
figure 1

Pathophysiology of acute kidney injury following intravascular contrast administration

Full size image

Volume and type of contrast agent

Several studies have identified an association of a higher volume of iodinated contrast with increased risk of AKI [12]. While such analyses are confounded by the recognition that sicker patients with more complex clinical presentations (who are hence at higher baseline risk for AKI) may undergo procedures that require higher volumes of contrast, this ‘dose-response’ relationship between contrast volume and risk of renal injury supports the nephrotoxic potential of these agents. Furthermore, past studies that compared the effects of different contrast agents support their causal association with AKI [13]. The initial contrast media used in clinical practice were ‘high osmolal’ with osmolalities several fold greater than blood (i.e., 1500–2000 mOsm/kg). Following the introduction of ‘low-osmolal’ contrast media (osmolality ~ 600–850 mOsm/kg), clinical trials and meta-analyses demonstrated lower risk for CA-AKI with these agents compared with ‘high-osmolal’ media [14]. A differential risk of AKI was also observed in certain more recent studies that demonstrated lower rates of CA-AKI following procedures that used iso-osmolal iodixanol compared with certain ‘low-osmolal’ agents [15]. While differences between iodixanol and ‘low-osmolal’ contrast regarding the risk for AKI were not shown in all prior trials, one would not expect to see any differences in the incidence of renal injury with these agents if they had no adverse effects on the kidneys.

Recent studies questioning the existence of CA-AKI

Notwithstanding past research that documented the pathophysiologic effects of iodinated contrast on the kidneys and differential risk of AKI based on volume and type of contrast agent, multiple recent studies have questioned the existence of CA-AKI [2,3,4,5,6]. A meta-analysis by McDonald and colleagues that included 13 studies with a total of 25,950 patients demonstrated that the risk of AKI following procedures with intravascular contrast administration was similar to the risk following procedures that did not utilize contrast (relative risk, 0.79; 95% CI, 0.62–1.02) [3]. The authors also reported no differences in the need for dialysis or death based on the receipt of contrast. More recently, Wilhelm-Leen et al. compared the incidence of AKI in a large cohort of hospitalized patients who did and did not undergo contrast-enhanced procedures [7]. In adjusted analyses, the incidence of AKI was 5.1% in patients who received contrast compared with 5.6% in those who did not (adjusted odds ratio, 0.93; 95% CI, 0.88–0.97).

While these and several other studies form the basis for the current hypothesis that CA-AKI does not exist, careful inspection of these studies demonstrates certain methodologic limitations that raise questions about the findings. First, all of these studies were retrospective observational analyses that relied on data that had been collected as part of routine clinical care. As such, the results were based solely on those patients in whom renal function was assessed prior to and following radiographic procedures. Furthermore, differential assessment of kidney function, regardless of reason, could not be fully accounted for in the analyses. Second, the use of preventive care to mitigate the risk of CA-AKI could not be fully evaluated. For example, demonstrating that the incidence of AKI is similar in patients who receive aggressive intravascular volume expansion before receiving iodinated contrast compared with patients who undergo non-contrast enhanced procedures would not establish that intravascular contrast is not nephrotoxic. Finally and most importantly, these studies could not fully account for factors that influenced providers’ decisions regarding the use of intravascular contrast. Patients at higher baseline risk of AKI were almost certainly less likely to receive intravascular contrast than patients at low baseline risk. No degree of statistical adjustment or propensity score matching can fully account for all potential confounders or eliminate the effect of indication bias. In fact, the likelihood that there were substantial differences in baseline risks for AKI between patients who did and did not receive contrast is borne out in the findings of some of these studies that demonstrated statistically significantly lower rates of CA-AKI among patients who received contrast compared with patients who did not [4, 7]. Unless one believes that intravascular iodinated contrast is nephroprotective, this observation highlights the likelihood of confounding by indication in such studies.

Conclusion

In summary, prior research has elucidated pathophysiologic effects of iodinated contrast on the kidneys in animal models and humans, while studies documenting associations of volume and osmolality of contrast media with risk for renal injury support their nephrotoxic potential. Recent studies questioning the existence of CA-AKI have important methodologic limitations that confound interpretation of their findings. While secular trends including the use of lower volumes of less nephrotoxic contrast along with the widespread use of preventive care including intravascular volume expansion have likely contributed to decreased rates of CA-AKI and rendered severe renal injury a relatively rare complication of contrast administration alone, these factors have not eliminated the existence of this iatrogenic condition. Continued vigilance and appropriation of evidence-based preventive care in the highest risk patients remains essential.