A 73-year-old man with stage III chronic kidney disease (CKD) presents to the emergency department with acute left–upper quadrant pain. Serum creatinine is 2.1mg/dL (eGFR 30 mL/min). Noncontrast computed tomography of the abdomen identifies small bowel inflammation and extensive atherosclerosis. Acute mesenteric ischemia is suspected, but further characterization requires intravenous contrast–enhanced images. He and his family worry about the safety of IV contrast and ask to speak with you.
Intravenous iodinated contrast material enhances tissue conspicuity in CT imaging and improves its diagnostic performance. Several case reports published in the 1950s suggested that IV administration of high-osmolality contrast provoked acute kidney injury. An ensuing series of studies associated contrast utilization with renal impairment and additional data extrapolated from cardiology arteriography studies further amplified these concerns.
Contrast media use is often cited as a leading cause of hospital-acquired acute kidney injury.1 The associated fear of causing renal impairment or provoking the need for dialysis frequently leads clinicians to forgo contrast-enhanced CT studies or settle for suboptimal noncontrast imaging even in situations where these tests are clearly indicated. The potential for inadequate imaging to contribute to incomplete, delayed, or incorrect diagnoses represents an ongoing patient safety issue.
A growing body of literature suggests the risks of contrast-associated acute kidney injury are overstated, implying the truer danger lies with inadequate imaging, not contrast media utilization. This review discusses the definitions, risks, and incidence of contrast-associated acute kidney injury, informed by these recent studies.
Overview of the data
Definitions of contrast-induced renal dysfunction vary in clinical studies and range from a creatinine rise of 0.5-1 mg per deciliter or a 25%-50% increase from baseline within 2-5 days following contrast administration. In 2012, the Kidney Disease Improving Global Outcomes working group proposed the term “contrast-associated acute kidney injury” (CA-AKI) and defined it as a plasma creatinine rise of 0.3 mg/dL within 48 hours of contrast exposure, a creatinine increase by a factor of 1.5 over baseline within 7 days of contrast administration, or a urinary volume less than 0.5 mg per kg of body weight within 6 hours of contrast exposure (AKI Network or “AKIN” criteria for CA-AKI).2 Owing in part to inconsistent definitions and partly because of multiple potential confounders, the true incidence of contrast-associated acute kidney injury is uncertain.
The pathogenesis of CA-AKI is incompletely understood, but proposed mechanisms include direct tubular cytotoxic effects; reductions in intrarenal blood flow from contrast material–provoked arteriolar vasoconstriction and contrast-induced increases in blood viscosity; and renal microvascular thrombosis.
Risk factors for CA-AKI overlap with those for acute kidney injury in general. These include CKD, concurrent nephrotoxic medication use, advancing age, diabetes, hemodynamic disturbances to include intravascular volume depletion, systemic illness, and rapid arterial delivery of a large contrast volume.
Current American College of Radiology guidelines state that intravenous isotonic crystalloid volume expansion prior to contrast administration may provide some renal protection, although randomized clinical trial results are inconsistent. The largest clinical trials of N-acetylcysteine showed rates of CA-AKI, need for dialysis, and mortality were no different than placebo. Studies of intravenous sodium bicarbonate show outcomes similar to normal saline.
Introduced in the 1950s and used until the early 2000s, the osmolality of high-osmolality contrast material (HOCM) is roughly five times that of blood (1551 mOsm/kg H2O).3 The early case reports first identifying concern for contrast-induced renal damage were of HOCM used in angiography and pyelography testing. Multiple follow up clinical studies measured creatinine levels before and after contrast administration and classified the percentage of patients whose creatinine level rose above an arbitrary definition of renal injury as having contrast-induced renal injury. These studies formed the basis of the now longstanding concerns about contrast-associated renal dysfunction. Importantly, very few of these HOCM studies included a control group.
Following multiple studies demonstrating an improved safety profile with a similar image quality, the Food and Drug Administration approved low-osmolality contrast (LOCM, 413-796mOsm/kg H2O) in 1985. Early adoption was slow because of its significantly higher cost and incomplete Medicare reimbursement. Prices fell following generic LOCM introduction in 1995 and in 2005 Medicare approved universal reimbursement, leading to widespread use. The FDA approved an iso-osmolality contrast material (290 mOsm/kg H2O) in the mid-1990s; its safety profile and image quality is similar to LOCM. Both LOCM and iso-osmolality contrast material are used in CTs today. Iso-osmolality contrast is more viscous than LOCM and is currently more expensive. Iso-osmolality and LOCM have similar rates of CA-AKI.
A clinical series published in 2008 examined serum creatinine level variation over 5 consecutive days in 30,000 predominantly hospitalized patients who did not receive intravenous contrast material. Investigators simulated contrast administration between days 1 and 2, then observed creatinine changes over the subsequent days. The incidence of acute kidney injury following the simulated contrast dose closely resembled the rates identified in earlier studies that associated contrast exposure with renal injury.4 These results suggested that changes in renal function commonly attributed to contrast exposure may be because of other, concurrent, clinical factors.
A 2013 study compared 8,826 patients with stable renal function who received a low-osmolality contrast-enhanced CT with 8,826 patients who underwent a noncontrast study.5 After 1:1 propensity matching, they found higher rates of CA-AKI (as defined by AKIN criteria) among only those with baseline eGFR less than 30 mL/min. There was a trend towards higher rates of CA-AKI among those with baseline eGFR of 30-44 mL/min, and no difference among the bulk of patients with normal or near normal baseline renal function.
Another large propensity score–matched study published in 2014 compared 6,254 patients who underwent a contrast-enhanced CT with 6,254 patients who underwent a nonenhanced CT.
Investigators stratified this predominantly inpatient cohort by baseline eGFR. Results demonstrated similar rates of AKI between contrast material and non–contrast material cohorts. They concluded that intravenous contrast administration did not significantly affect the risk of acute kidney injury, even in patients with impaired renal function. The authors noted that the difference in contrast-mediated nephrotoxic risk in patients with eGFRless than 30 between their study and the Davenport study could be explained by their use of a different definition of CA-AKI, differences in propensity score calculation, and by enrolling greater numbers of patients with impaired kidney function in their study.6
Finally, a large single-center study published in 2017 included 16,801 ED patients divided into three groups; patients who received a contrast-enhanced CT, patients who underwent a noncontrast CT study, and a set of patients who did not undergo any CT imaging. Patients with creatinine levels under .4 mg/dL or over 4 mg/dL were excluded from initial analysis.
Investigators stratified each patient group by serum creatinine and eGFR and utilized both traditional contrast-induced nephropathy (serum creatinine increase of .5 mg/dL or a 25% increase over baseline serum creatinine level at 48-72 hours) and AKIN criteria to evaluate for acute kidney injury. Propensity score analyses comparing the contrast-enhanced group and two control groups failed to identify any significant change in AKI incidence. The authors concluded that, in situations where contrast-enhanced CT is indicated to avoid missing or delaying potential diagnoses, the risks of diagnostic failure outweigh any potential risks of contrast induced renal injury.7
While these three studies utilized control groups and propensity score matching, they are retrospective in nature and unknown or omitted confounding variables could be present. Together, though, they contribute to a growing body of literature suggesting that the risk of contrast-associated AKI relates less to the contrast itself and more to concurrent clinical factors affecting kidney function. Ethical concerns have to date prevented the conduct of a randomized trial of IV contrast in CT scanning. Table 1 summarizes the findings of these three studies.