Biomarkers in acute kidney injury due to contrast induced nephropathy
Background: Despite preventive guidelines, iatrogenic contrast-induced nephropathy (CIN) ranks third as a cause of hospital acquired acute kidney injury (AKI), and impacts significantly on morbidity and mortality and is associated with high hospital costs. In Sub-Saharan Africa, the rates and risk factors for CIN remain unexplored. Despite the positive association of genetic polymorphisms in the TNFα and IL10 genes with CIN in Asian populations, the CIN genetic susceptibility in other races is unknown. Serum creatinine is a sub-optimal biomarker for the early diagnosis of CIN resulting in delayed interventions. This study investigated rates, risk factors and outcomes of CIN, the influence of genetic susceptibility to CIN in the black population and lastly, the accuracy of novel biomarkers in the early diagnosis of CIN and prognosticating patient outcomes. Methods: This was a prospective case-controlled study conducted at Charlotte Maxeke Johannesburg Academic Hospital, in South Africa from January 1, 2014 to December 30, 2015.Hospitalized patients undergoing enhanced computed tomography and angiography were consecutively recruited to the study and followed up for development of CIN. CIN was defined as an increase in serum creatinine >25% or an absolute increase of >44 μmol/l from baseline at 48-72 hours after exposure to contrast media. In the second part of the study, a nested case-controlled cohort that included 30 CIN patients and 60 controls (those undergoing contrast administrations and not meeting CIN criteria) were ethnically matched for gender, and age in a case: control ratio of 1:2 at all-time intervals. Sera for neutrophil gelatinaseassociated lipocalin-2 (NGAL), cystatin C, beta-2 microglobulin (β2M), interleukin 18 (IL18), IL10, and tumor necrosis factor alpha (TNFα) were collected at four time points: baseline (pre-contrast), 24 hours, 48 hours and ≥5-7 days after contrast administration and their concentrations were determined using luminex assays and an enzyme linked immunosorbent assay for β2M as per manufacturer’s instructions. The areas under receiver operating characteristic curves (AUROC) were generated to determine accuracy of novel biomarkers to diagnose CIN and CIN mortality. Genomic DNA was extracted from peripheral blood samples of 208 black South Africans using the Maxwell DNA purification kit (Promega AS1010, USA) and their genotypes for - 308(rs1800629) and -857(rs1799724) in the TNFα gene and -592(rs1800872), - 819(rs1800871), -1082 (rs1800896) and +1582(rs1554286) in the IL10 gene were determined by restriction fragment length polymorphism (RFLP). Results: We recruited 371 hospitalized patients (mean age 49.3±15.9); the rates of CIN were4.6% and 16.4% respectively, using an absolute or relative increase in serum creatinine from baseline. Anaemia was an independent predictor for the development of CIN (RR 1.71, 95% 1.01-2.87; p=0.04). The median serum albumin was 34 g/l (IQR: 29-39.5) vs. 38 g/l (IQR: 31-42), p=0.01 in the CIN and control groups respectively.Mortality was significantly increased in the CIN group (22.4% vs. 6.8%; p<0.001), and CIN together with anaemia predicted mortality with a 2-fold (p=0.01) and a 3-fold (RR p=0.003) riskrespectively. The median cystatin C at 24 hours (p<0.001) and β2M(at all-time points)levels were significantly higher in the CIN group compared to controls. The median cystatin C at 24 hours and β2Mlevels at 48 hours were 856.59 ng/ml (IQR 620.75-1002.96) vs. 617.42 ng/ml (IQR 533.11-805.20); p<0.001 and 5.3 μg/ml (IQR 3.8-6.9) vs. 3.3 μg/ml (IQR 2.7-4.5); p<0.001 with AUROCs of 0.75 and 0.78 respectively for early CIN discrimination.Pre-contrast IL18 (p <0.001), β2M (p=0.04) and TNFα (p<0.001) levels were significantly higher in the nonsurviving group and their AUROC were 0.83, 0.82 and 0.94 for CIN+ mortality. Baseline NGAL was a better marker for excluding patients at higher risk of developing CIN with negative predictive and positive predictive values of 0.81 and 0.50 respectively. The frequency of TNFα -308 AA genotype was significantly increased in the CIN group compared to controls (13.3% vs.1.82%, p=0.016) and the presence of the TNFα-308 AA (high producer) vs. GA genotypes was associated with a 9-fold CIN risk (9.24, 95% CI, 1.88-45, p=0.006). The IL10-1082 AA-allele (low producer) was significantly higher in the non-surviving CIN+ patients compared to controls (p=0.01). Conclusions:CIN occurred at a relatively high rate in our study and predicted poorer clinical outcomes. The presence of CIN and anaemia positively predicted mortality. Caution should be exercised in patients with anaemia and hypoalbuminaemia undergoing contrast studies. Serumcystatin C was the best novel biomarker for the early diagnosis of CIN and while baseline NGAL is superior as a biomarker for excluding patients at higher risk for CIN. IL18, β2M and TNFα are the best novel biomarkers for predicting the prognosis of patients with CIN. Increased frequency of the TNFα-308 AA genotype is a predisposing factor for CIN development. The low producer IL10-1082 AA genotype decreases survival in patient with CIN.
A thesis submitted to the Faculty of Health Sciences, University of the Witwatersrand, in fulfilment of the requirements for the degree of Doctor of Philosophy Johannesburg, 2016