Development of a Biomarker Panel for the Earlier Prediction of Acute Kidney Injury in Patients With Diabetes Mellitus Undergoing Coronary Revascularisation

Diabetes mellitus is an important risk factor for the development of contrast induced nephropathy (CIN), acting as a 'risk multiplier', amplifying the risk of acute kidney injury in these patients. There are important prognostic implications following the development of CIN and it is associated with a significantly increased mortality at 1 year. Diabetes is a major risk factor for coronary disease and these patients often have significant co-morbidities.

Currently creatinine is used to assess risk but this often lags behind clinical status. There is a pressing need for the development of novel, specific biomarkers to improve the detection and treatment of CIN and improve patient outcome in this high risk population.

This is a single centre,study in diabetic patients already undergoing a planned procedure, that is, a percutaneous coronary intervention (PCI). They are patients who are deemed to have an enhanced risk of contrast induced nephropathy by virtue of their diabetic and renal status, the latter being defined by a reduced eGFR which is a marker of renal disease and is based on the creatinine and characteristics of the patient. No additional interventions that are not part of their routine clinical care will be undertaken in these patients. We will be identifying natural biomarkers by obtaining serum and urine samples from these patients.

From a retrospective audit of the cardiac catheter lab database and a review of the literature we have estimated that a sample size of approximately 250 patients with DM and CKD (eGFR < 60 ml/ml) will be needed. We envisage that that we will encounter at least 50 cases of CIN from this cohort. (based on an expected incidence of CIN between 15-30% in this group). Looking for a study rate difference of at least 25%, for power of 95% and confidence intervals of 95% (with Fleiss correction) we will need at least 204 evaluable patients (to avoid Type 2 error). In view of potential drop-out of 10-15% we therefore intend to recruit 250 patients By using C statistics (Receiver operator curve analysis) we will be able to confirm or otherwise that either a particular biomarker or a combination of several biomarkers within 18 hours of procedure will increase the predictive power of CIN developing 72 hours later.

As part of their normal care patients will arrive in hospital on the morning of their planned PCI. They will at some point during the day undergo their PCI. Blood and urine will be taken just prior to the procedure and then at 2 hours, 4 hours, 8 to 12 hours, pre discharge and 3 days after the PCI.