RESULTS
As expected, the rabbits from DM group had a higher median venous plasma glucose level [25.6 (IQR 18.4-30.5) vs 6.0 (IQR 5.6-6.5) mmol/l, p<0.001]. In the DM group, glucose concentration and DM duration had weak significant correlation with each other (r = 0.308, p = 0.025).
First, we tested, if DM duration and/or plasma glucose level had any associations with arrhythmia inducibility. Diabetic conditions were associated with arrhythmogenesis during reperfusion but not ischemia. Table 1 shows that the diabetic animals did not differ from the controls concerning ischemic arrhythmias, but the number of arrhythmias induced at reperfusion was higher in the DM group. This increase in reperfusion arrhythmogenicity developed only in the prolonged DM group (Table 1). In the diabetic animals, both DM duration and glucose concentration were associated with reperfusion VT/VF in univariate logistic regression analysis (OR 1.058; 95% CI 1.025-1.092; p < 0.001; and OR 1,119; 95% CI 1,045-1,198; p = 0.001; respectively). In multivariate logistic regression analysis, only DM duration remained an independent predictor of reperfusion VT/VF (OR 1.060; 95% CI 1.006‑1.117; p = 0.029). No associations were found between ischemic VT/VF on one hand and neither DM duration, nor glucose concentration on the other hand.
Then, to find out which electrophysiological parameters were affected by DM duration, we compared electrophysiological parameters in the control group, and groups with short and prolonged DM (Fig. 1). The groups did not differ in DOR, time of epicardial breakthrough, duration of epicardial and total ventricular activation. Among the studied electrophysiological variables obtained by ventricular epicardial contact potential mapping, only duration of ARIs differed between the groups being increased in the prolonged DM group.
Within the diabetic group (any DM duration), we tested which electrophysiological mapping parameters were associated with DM duration and/or glucose concentration in multivariate linear regression analysis (Table 2). Duration of epicardial activation was associated with glucose concentration, whereas total ventricular activation and ARI duration were associated with DM duration. Epicardial isochronal maps show prolongation of ARIs particularly in the apical areas of ventricular epicardium in the animals with DM, especially at the long follow-up (Fig. 2). DOR demonstrated no associations with neither DM duration, nor glucose concentration.
We established electrophysiological determinants of ischemic and reperfusion VT/VF development. Table 3 shows results of univariate logistic regression analysis of arrhythmia predictors. It demonstrates association of VT/VF development during ischemia with activation time of epicardial breakthrough, DOR and average ARI duration. On the other hand, reperfusion VT/VFs were associated with only average ARI duration. ROC curve analysis demonstrated significant association of prolonged ARI with reperfusion VT/VF inducibility (AUC 0.781, p = 0.001). The optimal cut-off for the ARI ≥ 120 ms predicted reperfusion VT/VF with sensitivity 0.80 and specificity 0.75 (OR 8.119 95%CI 2.334-28.247, p = 0.001 in logistic regression analysis).