Patients with cancer rely on contrast-enhanced imaging for the diagnosis and staging of disease, monitoring of response to anticancer treatments, and monitoring of the risk of disease progression or recurrence during follow-up. Contrast-associated acute kidney injury (CA-AKI) occurs as a complication of diagnostic and interventional procedures. AKI is associated with long-term renal dysfunction, increased mortality, prolonged hospitalization, increased health care costs, and increased risk of adverse clinical outcomes in patients with cancer. Therefore, its prediction and prevention are important topics for many physicians, including radiologists, oncologists, and nephrologists. Literature on the evaluation of CA-AKI after contrast-enhanced CT in patients with cancer is limited [1]. Patients with cancer may additionally have unique risk factors for AKI, including exposure to tumor cytokines, nephrotoxic chemotherapy, and complications of cancer treatment.

Prediction models are crucial for risk stratification [2, 3]. In this large retrospective study, the authors developed a prediction model using simple clinical variables (hematologic malignancy, diuretic use, use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, chronic kidney disease stage 3a or greater, low serum albumin level, low platelet count, proteinuria, diabetes mellitus, heart failure, and higher volume of contrast medium) to estimate risk of CA-AKI in individual patients with cancer undergoing contrast-enhanced CT. The model showed good discrimination and calibration. The proposed risk score can be easily calculated at the bedside and integrated into the EMR. This model is easily applicable to clinical practice and is reasonably accurate. It can help assess benefits and risks of contrast-enhanced CT in patients with cancer. Such a model for predicting which patients are at higher risk of CA-AKI would facilitate timely implementation of preventive strategies and improve overall patient outcomes. Thorough validation and further large prospective studies are needed to confirm the generalizability of the model and its impact on the incidence of CA-AKI in patients with cancer and on the morbidity and mortality associated with CA-AKI.