AKI and Renal Parenchyma Attenuation in Hospitalized Patients With COVID-19

ABSTRACT

BACKGROUND AND AIMS Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) had a worldwide spread since early 2020 and a lot of studies concerning the diagnostic and prognostic role of chest computed tomography (CT) on coronavirus disease (COVID-19) has been published. Renal involvement might be present in up to 75% of cases, significantly impacting on the prognosis. The aim of this study is to clarify the role of opportunistic kidney assessment on non-enhanced chest CT and to evaluate if radiological findings could be associated with relevant clinical information regarding kidney function and patient's prognosis in hospitalized patients with COVID-19. METHOD We collected data on patient demographics, comorbidities, chronic medications, vital signs, baseline laboratory test results and in-hospital treatment in patients with COVID-19 consecutively admitted to our Institution who underwent chest CT. The standard chest CT-scan acquired in full inspiration include large part of both kidneys as per protocol. Three regions of interest (ROI) of 0.5–0.7 cm2 were positioned in every kidney, right and left to include both the cortex and the medulla. The mean values of attenuation of kidney regions were analysed. The primary and secondary outcomes were the occurrence of acute kidney injury (AKI), in-hospital and 9 months of death for all causes. RESULTS A total of 86 patients with COVID-19 and unenhanced chest CT were analyzed splitting the cohort into CT renal parenchyma attenuation (RPA) quartiles. Patients with a CT RPA below 24 Hounsfield unit (HU) were more likely to develop AKI when compared with other patients (×2 = 2.77, P = .014). An AKI-specific cut-off point of RPA was identified by performing a survival receiver operating characteristic (ROC) curve. At multivariate logistic regression analysis, being in the first quartile of CT RPA was associated with a four-times higher risk of AKI (Table 1) after adjustment for age, gender, hypertension, kidney function at admission and other comorbidities. During a mean 22 ± 15 days of admission, 32 patients died (37.2%). Patients with lower values of RPA at CT (first quartile, <24 HU) were not at a higher risk of death compared with patients with RPA ≥ 24 HU, as shown by Kaplan Maier curve (Fig. 1) and by multivariate Cox regression analysis [HR 1.84 (95% CI 0.82–4.13);P = .14].Figure 1 Data patients, grouped by AKI situation (without AKI, AKI on arrival or AKI during admission). CONCLUSION The association between AKI and RPA < 24 HU was independent of age, gender, creatinine and comorbidities. RPA values seemed to be predictive of AKI development in COVID-19 patients who underwent chest CT, suggesting RPA values could significantly improve patients’ care. The opportunistic measure of RPA could help physicians identifying patients with a higher risk of AKI, and this increased awareness could guide choices for diagnostic and therapeutic procedures.