Abstract
Acute Kidney Injury (AKI) can affect one in five major abdominal surgery (MAS) patients. AKI is defined by the ‘Kidney Disease: Improving Global Outcomes’ criteria, based on rises in serum creatinine and low urine output. This approach has limitations that significantly delay identification.
Novel biomarker tests have been investigated that aim to identify AKI rapidly, performing strongly in cardiac, transplant and vascular surgery where AKI rates approach 50%. A systematic review conducted to inform this thesis identified six papers examining biomarkers in MAS: methodological concerns were common throughout. Biomarker utility in MAS is thus unknown.
This observational study investigated the diagnostic performance of urinary biomarkers in 488 MAS patients. Biomarkers scrutinised were Neutrophil Gelatinase associated Lipocalin (NGAL), Insulin like Growth Factor Binding Protein-7 combined with Tissue Inhibitor of Metalloproteinases-2 (NC), Kidney Injury Molecule-1(KIM-1) and Dickkopf-3 (DKK-3). Concentrations were measured four hours post-operatively and assessed using area under the receiver operator curve analysis (AUROC).
AKI by serum creatinine (AKISCr) occurred in 8.8% of patients, and by urine output (AKIUO) in 49.6%. NGAL had the highest AUROC for AKISCr (0.74), followed by KIM-1(0.68) and NC (0.66). Performance improved with worsening AKI. AUROC values were below 0.65 for nearly all stages of AKIUO.
Logistic regression analysis revealed increasing age, body mass Index, operation duration, American Association of Anaesthesiologists grade, and male sex to be strongly associated with the development of AKISCr. A risk prediction model incorporating these factors was created and returned an AUROC of 0.8. With the best performing biomarker added (NGAL), the AUROC was 0.82.
This is the largest study interrogating AKI biomarkers in MAS, identifying a biomarker test and risk prediction model that perform well for AKISCr. An interventional study has been proposed that randomises biomarker positive MAS patients to an AKI ‘bundle’ or standard care. This may demonstrate whether early AKI identification with biomarkers translates into improved outcomes.