Key Considerations regarding the Renal Risks of Iodinated Contrast Media: The Nephrologist’s Role

The possibility that iodinated contrast media might cause acute kidney injury (AKI) was first raised in a 1954 case report of a multiple myeloma patient who developed anuria following intravenous pyelography with the high-osmolar agent Diodone [1, 2]. Many publications describing AKI after excretory urography, computed tomography, and noncoronary angiography soon followed, but the term contrast-induced nephropathy (CIN) was coined in the 1980s as a larger series of cases following coronary angiography were reported [3‒5]. Risk factors for CIN such as pre-existing kidney dysfunction, diabetes mellitus, hypotension, nephrotoxic drugs, diuretics, and intra-aortic balloon pump use were identified, and CIN was labelled as one of the most common causes of hospital-acquired AKI, contributing significantly to incident chronic kidney disease (CKD), end-stage kidney disease, and death [4, 6‒8]. With more frequent studies, the criteria adopted to define CIN narrowed. In 1999, the Contrast Media Safety Committee of the European Society of Urogenital Radiology (ESUR) suggested an increase in serum creatinine >25% or 44 μmol/L (0.5 mg/dL) within 3 days of contrast administration in the absence of other causes of AKI-indicated CIN – a definition that was still in place in the 2011 ESUR guidelines [4, 9, 10]. In 2012, the Kidney Disease Improving Global Outcomes (KDIGO) Working Group adopted the term contrast-induced AKI (CI-AKI), defining AKI as a ≥50% increase in creatinine from baseline within 7 days of exposure, a ≥0.3 mg/dL increase within 48 h, or urine output reduced to ≤0.5 mL/kg/h for at least 6 h [11]. While CI-AKI is now the accepted terminology to describe kidney injury resulting from iodinated contrast media administration, post-contrast AKI (PC-AKI) and, more recently, contrast-associated AKI (CA-AKI) are used to distinguish a correlative rather than causative diagnosis since it is not possible to exclude other causes of AKI in many clinical and most research settings [6, 12, 13].

Until relatively recently, it was generally thought that iodinated contrast, whether administered intra-arterially or intravenously, was capable of causing AKI in high-risk patients [5]. Over the past decade, however, the clinical relevance of CI-AKI, or indeed whether it exists at all, has been called into question [5, 6, 14]. The sceptics have pointed to observational studies from large administrative healthcare databases reporting similar rates of AKI in patients undergoing contrast-enhanced computed tomography and those undergoing unenhanced scans [5, 6, 14‒17]. While early, uncontrolled studies may have overstated the risk of CI-AKI, these propensity score studies are not without methodological issues [5]. Despite sophisticated matching techniques, selection bias cannot be completely excluded because it lacks randomisation. Higher risk patients are less likely to receive contrast, biasing the studies towards the null [5, 14, 18]. Another limitation is that patients with severely decreased estimated glomerular filtration rate (eGFR), at highest risk of CI-AKI, were underrepresented in these studies [5, 6]. The debate is important because it has clinical consequences for patients with CKD. If the risk is overstated, fear of CI-AKI may inhibit a clinician from ordering or performing the diagnostic or therapeutic contrast-enhanced imaging necessary to achieve maximal benefit for the patient [5, 6, 14]. Chertow and colleagues [19], for example, found coronary angiography to be underutilised in patients with CKD presenting with myocardial infarction, that was associated with significantly higher 1-year mortality [19]. Alternatively, if there is less focus on the kidneys and CI-AKI risk is underestimated, patients may be unnecessarily exposed to a nephrotoxic agent capable of causing AKI that may have both short- and long-term clinical implications [5]. Although the true incidence of CI-AKI is not clear, it remains a real phenomenon, and a high contrast dose, arterial administration, and impaired kidney function, especially in combination, remain strong risk factors for this complication [13, 18].

Different disciplines are involved in the decisions and delivery of contrast-based imaging, and the settings, indications, and patient profiles contribute to the complexity. For instance, emergent evaluations for pulmonary embolism in the emergency room or coronary angiography for a myocardial injury are quite different from elective procedures for evaluating a stable patient with a symptomatic abdominal mass or from serial studies to track the progression of a tumour size following chemotherapy. The multidisciplinary nature of these diagnostic procedures contributes to variations in care and complexity and the perceived prevalence of CI or CA-AKI. Renal complications following contrast administration remain of great concern since they are associated with worse patient outcomes, including accelerated progression of underlying CKD and increased mortality [20, 21]. CA-AKI may also be on the increase. Data from a US administrative database with over 2 million patients undergoing percutaneous coronary intervention (PCI) show that the incidence of CA-AKI increased from 6.0% at the beginning of 2012 to 8.4% at the end of 2017 overall, and from 18.0 to 28.4% in those with CKD over the same time period. The incremental cost associated with AKI in this study was estimated at $7,000–$9,000 per case [22].

Considering the high human and financial costs, CA-AKI prediction and avoidance are daily issues for the specialities ordering and performing contrast-enhanced procedures [23]. However, guidelines for contrast use are often not the same across different settings, and disciplines and testing for kidney function post-contrast exposure is uncommon. Recognition of CA-AKI may thus be delayed and often comes to light much later when a patient is seen by a nephrologist. Consequently, risk mitigation is important for those dealing with the consequences of CA-AKI because there are no effective therapeutics to reverse the injury once it occurs [24, 25]. Early contributions by nephrologists in the care of patients at high risk for CA-AKI sharing their knowledge about appropriate prevention could become a cornerstone of better management [26, 27].

The kidney is particularly susceptible to damage after contrast medium administration. After intravascular injection, a contrast medium is diluted in the bloodstream and immediately distributed throughout the extracellular fluid. Being poorly bound to serum albumin, it is freely filtered by glomeruli [28]. With 99% of the filtrate reabsorbed, the concentration of contrast rises 100-fold in the tubules [14, 29]. In patients with a reduced GFR, contrast remains in circulation for longer. With fewer working nephrons, each one has a larger load of contrast to filter, and the exposure burden per nephron is increased [25]. As a result, a lower volume of contrast may result in adverse effects in patients with pre-existing renal impairment, and their risk of AKI may be over 10-fold greater than in those with normal kidney function [30, 31]. Current guidelines broadly agree on three main considerations to protect the kidneys in patients undergoing contrast-enhanced procedures: the risk status of the individual, their level of hydration, and the contrast used [25, 32]. Pre-existing CKD is the strongest patient-related risk factor [21]. The risk of CA-AKI is directly correlated to the stage of CKD with rates of around 5%, 10%, 15%, and 30% when the eGFR values are ≥60, 45–59, 30–44, and <30 mL/min/1.73 m², respectively [33]. While not an independent risk factor, diabetes mellitus may amplify susceptibility in patients with CKD [21]. An adequate body hydration status leads to a lower concentration of contrast agent at the site of the renal tubules, which will reduce interaction of the contrast with the kidneys [20, 25]. Hydration is not without side effects, however, especially in patients with reduced cardiac and kidney function who are at risk of fluid overload [20]. While the choice of fluid has been debated, the results of the large, well conducted Prevention of Serious Adverse Events Following Angiography (PRESERVE) trial suggests intravenous normal saline should be the fluid of choice for patients at risk rather than sodium bicarbonate or acetylcysteine [21, 34, 35]. When the intravenous route is not feasible, guidelines recommend oral hydration still be encouraged for patients considered at high risk, even though studies suggesting equivalent benefit to the intravenous route are limited [32, 33]. The third consideration in mitigating risk is the contrast used for the procedure [25]. As the volume of contrast goes up, so too does the risk of AKI; for every 100 mL of contrast, the risk of developing post-procedural AKI can increase between 12 and 28% [36]. While there is not a “one size fits all” solution to how much contrast can be safely administered in high-risk patients, a general rule is that the volume used should not exceed two to three times the patient’s eGFR [37, 38] and that the gram iodine/eGFR ratio should not exceed 1.0 [39]. The type of contrast agent is very important [37]. Use of low-osmolar and isoosmolar agents significantly reduces the incidence of CI-AKI compared with high-osmolar contrast media [37]. However, the data regarding whether the risk is lower with isoosmolar contrast compared with low-osmolar contrast are conflicting with some studies finding no difference [30, 37, 40]. Three recent real-life studies may shed further light on this issue. Each used contemporary U.S. hospital data to retrospectively compare the cardiorenal impact of isoosmolar and low-osmolar contrast in patients at high risk for adverse renal events. The different studies considered those undergoing intra-arterial diagnostic or treatment procedures, those undergoing endovascular aneurysm repair (EVAR), and those undergoing peripheral vascular procedures. In each scenario, isoosmolar iodixanol was associated with a significant reduction in major adverse renal and cardiovascular events versus low-osmolar contrast media [41‒43].

Risk factors predisposing the development of CA-AKI can be considered patient-related (e.g., whether they have pre-existing comorbidities such as CKD and/or diabetes) and procedure-related risk (e.g., whether the procedure is arterial or venous) [23, 44]. For better risk stratification of patients, a series of models that incorporate both patient and procedural factors have been proposed [23, 45‒47]. While a strength of these risk-stratification models is that they are derived from data based on large numbers of patients, most were developed in the PCI setting. Consequently, they may not be readily generalisable and require recalibration if used with a different patient group [21, 48, 49]. Another limitation of some models is that they cannot be used for pre-procedure identification of at-risk patients because they include operational variables, such as contrast volume, which are usually not known until the procedure is executed [21, 45]. However, the adoption of these scores may reduce the tendency to undertreat patients who require procedures with contrast administration [45, 48, 50]. Formal risk stratification also supports the conversation between doctor and patient, helping to guide informed consent [45].

The current international guideline consensus, however, is that patients at risk of PC-AKI are those with impaired renal function, without other risk factor-based screening [12, 13, 48, 51, 52]. Not all radiological and nephrological societies, however, agree with the restriction of patient-related risk factors for prophylaxis against PC-AKI to renal risk factors [39]. Imprecision in estimating renal function is unavoidable. Because GFR cannot be measured directly, endogenous filtration markers such as creatinine or cystatin are used to estimate GFR through an algorithm (eGFR). Up to 25% of people will have estimates more than 30% above or below their GFR [53]. There is also wide variation in eGFR values depending on the equation used such that it can alter a patient’s CKD stage classification [54]. There will also be some emergency situations where contrast-enhanced imaging is indicated but estimation of GFR is not practical [55].

While intuitively giving as much isotonic fluid as possible to enhance washout of contrast through the kidney may seem desirable, it may have a deleterious effect, particularly in patients whose cardiac function is impaired, increasing their risk of AKI as well as worsening heart failure [56, 57]. Novel ways to guide intravenous fluid administration have therefore been tried, including the use of haemodynamic parameters such as left ventricular end-diastolic pressure (LVEDP) and central venous pressure (CVP) [56, 58, 59]. The POSEIDON study investigated a fluid protocol involving LVEDP, a parameter routinely obtained during cardiac catheterisation which provides a measure of intravascular volume status. Compared with patients randomised to standard care, patients in the LVEDP group received over twice the volume of normal saline, and their risk of post-contrast AKI was reduced by over half [58]. The CVP-guided study had similar findings. Patients whose hydration infusion rate was dynamically adjusted according to CVP level every hour received 50% more isotonic saline, with rates of post-contrast AKI that were almost halved compared with those receiving standard hydration [59]. In settings where cardiac haemodynamic parameters are not readily available, bioimpedance vector analysis (BIVA) allows a non-invasive determination of hydration status. When used to identify patients with low total body fluid levels on admission, it allowed intravascular volume expansion to be adjusted, resulting in lower CA-AKI occurrence after angiographic procedures [60]. Other studies have investigated forced diuresis systems, where the delivery of intravenous fluid is matched to urine output [61, 62]. An automated system delivering saline plus furosemide achieved high urinary flow rates and was found to be superior to conventional hydration regimens in preventing post-contrast AKI in high-risk patients [61]. More recently, a non-automated matched hydration and forced diuresis protocol also proved effective in preventing post-contrast AKI in patients with CKD [62]. In patients undergoing interventional cardiology procedures, prophylactic administration of statins, in addition to hydration, appears to protect against CI-AKI, although this is less evident in patients with CKD [63].

Because a reduction in the amount of contrast administered is a mainstay of preventive therapy, it might be expected that patients at higher risk of AKI would be treated with less contrast [64]. Analysis of over 1.3 million patients included in the National Cardiovascular Data Registry (NCDR) CathPCI Registry, however, found little evidence that physicians were limiting the amount of contrast in patients at higher risk. There was substantial variation in contrast volumes used and in the incidence of post-contrast AKI among individuals performing PCIs, even after accounting for the complexity of the procedure or the degree to which patients were compromised, implying differences were due to physician practices [65]. Both the variation among physicians and the absence of an adjustment in contrast volume for patients at higher risk underscore an important opportunity to reduce post-contrast AKI [65]. A survey of international cardiologists suggests simply measuring contrast volume used would be a starting point for many; among the 506 individual respondents, some 40% did not accurately measure volume [66]. Using data from PCIs recently performed in 18 hospitals, researchers estimated that approximately one-quarter of patients undergoing PCI could have their CA-AKI risk meaningfully reduced by using the calculated safe contrast limit [67]. Several software tools are available to track contrast media volume administered to a patient, not just during one procedure but also cumulatively over time (e.g., when several examinations are performed within short period of time). Realising the benefits of such a clinical decision support system (CDSS) in practice is not without challenges. The Barnes-Jewish Hospital in St. Louis succeeded in navigating the challenges of CDSS implementation, reducing both contrast use and AKI incidence with a predictive model that estimated patient-specific risk and individualised safe contrast volume at the start of each case [68]. Similarly, James et al. [69] have shown the application of a CDSS for reducing AKI in patients undergoing PCI in a stepped wedge cluster randomised trial in Canada with a reduction in the incidence of AKI from 8.15 to 7.6% and an improvement in the application of IV fluids for hydration, and a reduction in contrast volume. Another way to optimise the amount of contrast medium is the use of devices such as DyeVert^TM, which are designed to reduce and facilitate monitoring of the volume of contrast medium dispensed during coronary or angiographic procedures while maintaining fluoroscopic image quality [70]. There is no evidence that extracorporeal therapies to remove contrast medium prevent the onset of AKI in patients with CKD [71].

Serum creatinine is a fundamental biomarker in CI-AKI, although with low sensitivity and specificity, but its increase is delayed after kidney injury (serum creatinine tends to remain relatively normal even in the presence of kidney damage, until approximately 50% of nephrons are lost) [72‒74]. Numerous other biomarkers have therefore been investigated for their predictive ability in CI-AKI, including cystatin C, urinary and serum neutrophil gelatinase-associated lipocalin (uNGAL and sNGAL), interleukin-18 (IL-18), kidney injury molecule 1 (KIM-1), and liver fatty acid-binding protein (LFABP) [72, 73, 75]. Some of these biomarkers represent changes in renal function and others reflect kidney damage [76]. Some seem better in certain situations or at certain timepoints – for example, not at baseline but after contrast exposure – making it difficult to draw clear conclusions on use in clinical practice [73, 75, 76]. Recently, the potential for urinary dickkopf-3 (DKK3), a stress-induced renal tubular epithelium-derived glycoprotein, in CI-AKI risk prediction was assessed in patients with CKD undergoing coronary and/or peripheral angiography and/or intervention [75]. Urinary DKK3/urine creatinine at baseline was more strongly associated with the risk of CI-AKI compared with uNGAL, sNGAL, and serum cystatin C; it was also the strongest predictor of persistent renal dysfunction [75]. Given that the toxic effects of contrast media are exhibited at the level of the renal tubule, it seems likely that urinary DKK3/urine creatinine is identifying patients whose renal tubules are stressed and therefore more susceptible to contrast-induced renal injury [77]. The identification of new biomarkers specific for CI-AKI needs to be coupled with the proposed consensus recommendations from Acute Disease Quality Initiative (ADQI) for identifying subclinical stages of AKI utilising a combination of kidney damage and functional markers applied together [72].

Because we have no effective therapies to treat AKI, prevention of post-contrast AKI is critical to improve outcomes [27]. The first step to implementing preventive measures is awareness. Whether through educational curricula, clinical meetings, or personal communication, nephrologists are well placed to promote knowledge of post-contrast AKI, including its institutional and financial burden [78]. Since neither minimising contrast volume nor optimising intravenous hydration are consistently applied in current clinical care, nephrologists can also play a vital role in educating colleagues about the necessity to establish risk mitigation strategies for patients at high risk of AKI [27, 79]. Being part of a multidisciplinary team where cardiologists and nephrologists consider aspects of each speciality would seem essential in the design, implementation, and monitoring of CDSSs, such as the Contrast-Reducing Injury Sustained by Kidneys (Contrast RISK) initiative being evaluated in Alberta, Canada [69, 80]. In the case of patients perceived to be at high risk, it is preferable that nephrologists are involved in early decision-making rather than when AKI has already occurred [27, 77, 81‒83]. Moreover, acute kidney insults should not be considered isolated episodes [27]. The capacity of the kidney to increase its level of operation in response to demand is known as the renal function reserve (RFR) [84]. It is important that non-renal specialists understand the recovery of renal function may be sustained by RFR despite nephron loss, even if GFR returns to normal. As RFR is lost with repeated renal injury, the kidney may be more susceptible to further insults, increasing the risk of clinically evident AKI even in the presence of a mild insult [84, 85]. A key issue is to track the course of patients at high risk following a contrast exposure with a check of serum creatinine within a week of the procedure and at further intervals as determined by clinical need. Nephrologists should add this to their recommendations to improve care coordination and identify patients who need follow-up.

Despite divergent attitudes regarding the relevance of post-contrast AKI in actual practice, prevention is clearly better than no cure. With the involvement of multiple specialities adding complexity, the identification of patients more susceptible to worse outcomes and implementation of tailored prevention and mitigation strategies are at best difficult. Equally, avoiding contrast to patients considered at high risk can lead to delays in diagnosis and them missing out on treatment which is ultimately more harmful than the contrast itself. Nephrologists are useful to support colleagues in finding the “equilibrium point” between over- and under-estimation of contrast media-related risks, with the potential to achieve enhanced outcomes for many high-risk patients.

This was a review of the published literature; no ethical approval was required.

The authors have no conflicts of interest to declare.

The authors did not receive any funding.

All authors were involved in the conception, drafting, critical revision, and approval of the manuscript.

All data presented has been published previously, as cited, and no new data were generated for this manuscript.