A Clinical, Proteomics, and Artificial Intelligence-Driven Model to Predict Acute Kidney Injury in Patients Undergoing Coronary Angiography

Clinical Cardiology

January 2019

Highlights & Key Findings

  • Acute kidney injury is a sudden decrease in kidney function with or without kidney damage, occurring over a few hours or days. 

  • To estimate diabetes- and nondiabetes-related acute kidney injury trends, CDC analyzed 2000–2014 data from the National Inpatient Sample (NIS) and the National Health Interview Survey (NHIS). Age-standardized rates of acute kidney injury hospitalizations increased by 139% (from 23.1 to 55.3 per 1,000 persons) among adults with diagnosed diabetes, and by 230% (from 3.5 to 11.7 per 1,000 persons) among those without diabetes.

  • The number of cardiac catheterization or Percutaneous Intervention (PCI) cases resulting in AKI rose almost 3‐fold from 2001 to 2011. Prevention strategies, such as the HART AKI test, are needed for at-risk patients to reduce future AKI risk. In this study, researchers assessed Prevencio’s AI-driven HART AKI diagnostic test on patients enrolled in Massachusetts General Hospital’s (MGH) Catheter Sampled Blood Archive in Cardiovascular Disease (CASABLANCA) study. The HART AKI test algorithmically assesses four variables (history of diabetes, blood urea nitrogen to creatinine ratio, C-reactive protein, and osteopontin) that have a positive association with AKI risk, and an additional two variables (CD5 antigen-like and Factor VII) that have a negative association with AKI risk.

  • Patients were referred for angiography for various acute and non-acute indications. Of the 1251 patients enrolled, patients were excluded if they: did not undergo a coronary angiogram: had a history of renal replacement therapy: had missing blood urea nitrogen or creatinine values; and had an insufficient quantity of sample. This left 889 patients undergoing coronary angiography with available blood samples.

  • Procedural AKI was defined as an abrupt reduction in kidney function with an absolute increase in serum creatinine of more than or equal to 0.3 mg/dL, a percentage increase in serum creatinine of ≥50%, or a reduction in urine output (documented oliguria of <0.5 mL/kg per hour for >6 hours) within 7 days after contrast exposure.

  • Forty-three (4.8%) patients developed procedural AKI. Those who developed procedural AKI were older (70 vs. 67 years of age, p=0.04) and more likely to have prevalent diabetes mellitus (41.9% vs. 23.5%, p=0.01) or CKD (20.9% vs. 10.4%, p=0.04) Those who developed procedural AKI also had lower left ventricular ejection fraction at baseline (50.0% vs. 56.6%, p=0.04) and a higher percentage of them were prescribed an angiotensin converting enzyme inhibitor (ACEi)/angiotensin receptor blocker (ARB) compared to those who did not develop AKI (72.1% vs. 53.6%, respectively, p=0.02).

  • HART AKI had an area under the receiver operating characteristic curve (AUC) of 0.82 (p<0.001) for predicting procedural AKI. The optimal score cut-off had 77% sensitivity, 75% specificity, and a negative predictive value of 98% for procedural AKI (Figure 1). An elevated score was predictive of procedural AKI in all subjects (odds ratio=9.87; p<0.001).

Conclusions

  • In a typical at-risk population undergoing coronary angiography for various acute and non-acute indications, we describe a clinical and proteomics-supported biomarker model with high accuracy for predicting procedural AKI in patients undergoing coronary angiography. The ability to predict AKI may allow for earlier interventions in at-risk patients to reduce future AKI risk

Scroll Down for the Clinical Cardiology Publication

The Clinical Cardiology Publication