HbA1c and Postprandial Glucose both Have Excellent Probability for Diagnosis of NODAT at both 12- or 52 Weeks Post Kidney Transplantation Compared to Fasting Glucose Alone

Abstract

Introduction: Glycated haemoglobin (HbA1c) can be utilised to diagnose diabetes mellitus in the general population, but guidance regarding its utility in the transplant population is limited and lacks validation. Beyond 3-months post-transplantation HbA1c may be useful to diagnose new-onset diabetes after transplantation (NODAT) and easier to perform than an oral glucose tolerance test (OGTT). We analysed the sensitivity of three diagnostic markers for diabetes (fasting glucose, postprandial glucose and HbA1c) for diagnosis of NODAT at 12- or 52-weeks post kidney transplantation. Methods: In this prospective analysis, 72 non-diabetic kidney transplant recipients received an OGTT at weeks 12 and 52 post-transplantation with concomitant HbA1c analysis. All patients received standard immunosuppression of basiliximab induction with maintenance tacrolimus, mycophenolate mofetil and corticosteroids. Patients were classed with NODAT in accordance to American Diabetes Association guidelines; fasting blood glucose ≥ 7.0 mmol/L, postprandial blood glucose ≥ 11.1 mmol/L or HbA1c ≥ 6.5%. Correlation between continuous variables was by Pearson's or Spearman's rank test for parametric and non-parametric data respectively. Receiver operator characteristics (ROC curve analyses) were performed to determine the probability of having NODAT. All statistics were performed using SPSS software, with a p value < 0.05 considered significant. Results: NODAT was diagnosed in 22.9% and 19.6% of recipients at 12- and 52 weeks post-transplant respectively. ROC curve analyses for probability of NODAT at 12-weeks were; 12-week fasting glucose (AUC=0.713, p=NS), 12-week HbA1c (AUC=0.991, p< 0.001) and 12-week postprandial glucose (AUC=0.984, p=0.001). ROC curve analyses for probability of NODAT at 52-weeks were; 52-week fasting glucose (AUC=0.599. p=NS), 52-week HbA1c (AUC=0.935, p=0.001) and 52-week postprandial glucose (AUC=0.796, p=0.023). Incorporating both time points for analysis HbA1c had the greatest probability of predicting NODAT with an AUC 0.921 (p< 0.001), with sensitivity 84.2%, specificity 100.0% and accuracy of 96.8%. This compared favourably to postprandial glucose (AUC=0.773, p< 0.001), whilst fasting glucose had poor probability (AUC=0.591, p=NS). Excluding patients diagnosed with NODAT by HbA1c alone at 12- or 52-weeks, HbA1c still demonstrated excellent sensitivity for probability of NODAT in the remaining patients with an AUC of 0.926 (p< 0.001). 12-week OGTT showed HbA1c correlated strongly with postprandial glucose (r=0.700, P< 0.001), but had only moderate correlation with fasting glucose (r=0.475, p=0.002). 52-week OGTT showed moderate correlation between HbA1c and fasting glucose (r=0.338, p=0.041) with only a trend towards correlation between HbA1c and postprandial glucose (r=0.307, p=0.064). Conclusion: HbA1c is comparable to postprandial glucose for diagnosis of NODAT beyond 3-months post-transplantation, both of which are superior to fasting glucose. Due to ease of use, HbA1c should be formally incorporated into diagnostic classification for NODAT in patients beyond 3-months post-transplantation to facilitate immediate clinical utilisation.