Radiographic Contrast Media and the Kidney

Abstract

AKI is a potential complication of intravascular iodinated contrast exposure. Contrast-associated AKI, which typically manifests as small and transient decrements in kidney function that develop within several days of contrast administration, is associated with serious adverse outcomes, including progressive kidney dysfunction and death. However, a causal link between the small increases in serum creatinine that characteristically occur with contrast-associated AKI and serious adverse outcomes remains unproven. This is important given mounting evidence that clinically indicated, potentially lifesaving radiographic procedures are underutilized in patients with CKD. This has been hypothesized to be related to provider concern about precipitating contrast-associated AKI. Intravascular gadolinium-based contrast, an alternative to iodinated contrast that is administered with magnetic resonance imaging, has also been linked with potential serious adverse events, notably the development of nephrogenic systemic fibrosis in patients with severe impairment in kidney function. Patients hospitalized in the intensive care unit frequently have clinical indications for diagnostic and therapeutic procedures that involve the intravascular administration of contrast media. Accordingly, critical care providers and others treating critically ill patients should possess a sound understanding of the risk factors for and incidence of such outcomes, the ability to perform evidence-based risk-benefit assessments regarding intravascular contrast administration, and knowledge of empirical data on the prevention of these iatrogenic complications.

Kidney Cell Cycle Arrest and Cardiac Biomarkers and Acute Kidney Injury Following Angiography: The Prevention of Serious Adverse Events Following Angiography (PRESERVE) Study

“…Our finding of association between preangiography cardiac biomarkers and risk of major adverse kidney events suggests that the risk of adverse outcomes is more likely attributable to underlying cardiovascular disease and patient comorbidities rather than CA-AKI. Our findings are relevant because registry-based studies 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 have documented lower utilization of angiography among patients with CKD than among patients without CKD, which may relate to physician concern for precipitating CA-AKI among those with CKD. However, such underutilization of potentially life-saving contrast enhanced procedures in high-risk patients may be potentially deleterious because the risk of poor outcomes exist even before contrast exposure and that contrast exposure does not significantly increase the risk of CA-AKI in most patients.…”

Contrast-Induced Encephalopathy: A Rare Complication in a Patient on Peritoneal Dialysis with Several Risk Factors

“…A reduction in renal perfusion caused by a direct effect of CM and toxic effects on the tubular cells are the main factors in the pathophysiology of renal damage after iodinated CM administration [22]. In addition, other mechanisms have been described to contribute to kidney damage [27]. Elevated viscosity of CM at tubular level increases resistance to flow in renal tubules; increased fluid viscosity at distal level fosters tubular plugging.…”

Comparative analysis of multiphase CT volumetric kidney segmentation: fine-tuning to domain adaptation

“…These factors can be hard to control to obtain a normal imaging. 12 Thus, only being trained on standard contrast-enhanced datasets, such distinct characteristics of tissue appearance often lead to model failure in real clinical use cases. In this work, we studied phase-agnostic training methodologies to develop deep learning segmentation models resilient to changes in contrast uptake for kidney segmentation, and all the models are externally validated on Mayo Clinic contrast and non-contrast datasets which include primarily cases with kidney abnormalities, lower kidney function, transplant cases, our findings demonstrate optimal performance for the mixed phase and phase agnostic models.…”