Cardiac RF ablation is a widely accepted procedure for treatment of ventricular tachycardia and atrial fibrillation. Late gadolinium enhancement (LGE) MRI can be used to assess acute and chronic RF ablation lesions. However, LGE-MRI has serious limitations. This imaging technique requires contrast injection and the visibility of lesions and appearance considerably change with time after the injection. Recently, native (non-contrast) T1-weighted (T1w) MRI was proposed to visualize RF lesions immediately (< 3 hours) after ablation^1,2. It was shown that lesion (enhancement) dimensions from native T1w images acquired immediately after ablation procedure are well correlated with pathology and LGE-MRI results for permanent lesions (3-month post-ablation scar). Performing MRI study immediately after RF ablation procedure may be not practical in some clinical settings. The main aim of this work was to study how the visibility and volume of RF lesions in native T1w MRI changes with time after ablation.

RF ablations of right and left ventricles of canines (n=2, weight=25-27 kg) were performed according to protocol approved by the local IACUC. Focal RF lesions were created using CoolFlex RF ablation catheter (St. Jude Medical Inc.) at 30-40 Watts for 60 seconds. The ablations were performed in an MRI suite equipped with a 3T whole-body MRI scanner (Verio, Siemens Healthcare, Erlangen, Germany). At the end of each ablation, animal was moved to the scanner iso-center. The time lapse between end of ablation and the animal in scanner iso-center was less than 5 minutes. Imaging protocol included T1 and T2 mapping, double inversion recovery (DIR) prepared T2w TSE and HASTE, 3D T1w, followed by Gd-BOPTA injection (0.15 mmol/kg, Bracco Diagnostic Inc., Princeton, NJ) and 3D LGE. Imaging studies were repeated for all animals at 2 and 7 days after ablation. Native T1w images of whole heart were acquired using 3D respiratory navigated, saturation recovery prepared GRE pulse sequence. Typical scan parameters were TR/TE=3.1/1.4 ms, flip angle of 10 degrees, TI=400 ms, voxel size=1.25x1.25x2.5 mm. Fat saturation was applied immediately before data acquisition which was limited to 10% of RR interval.

Ablation lesions (enhanced regions) were manually segmented on native T1w images acquired at different time points. Lesion volume was calculated for each ablation (n=9). Lesion volumes for 2 and 7 days post-ablation studies were normalized by the corresponding lesion volumes from acute (0 day) study. The visibility of ablation lesions in T1w images was quantitatively assessed using Image Intensity Ratio (IIR) metrics. IIR was calculated as a ratio of mean signal of lesion and mean signal of normal myocardium. Area of normal myocardium was selected using T1w, T2w TSE, and LGE images to exclude edematous regions.

Representative T1w images of RF ablation lesions acquired acutely, 2, and 7 days post-ablation are shown in Figure 1. These images clearly demonstrate that visibility of ablation lesions in native T1w images drastically reduces few days after ablation. Quantitative analysis shows that both IIR and normalized lesion volume, decrease significantly (p<0.05) as early as 2 days post-ablation (Fig. 2 and Fig. 3). This trend continues further with time post-ablation, resulting in the inability to detect many ablation lesions on native T1w images acquired 7 days post-ablation.Our results demonstrate that reduction of T1 relaxation time of RF ablated myocardium has transient nature. The visibility of RF ablation lesions and volume of the corresponding enhanced regions in native T1w MRI drastically reduce with time after ablation. Non-contrast T1w MRI should be performed as soon after RF ablation procedure as possible to achieve high contrast between ablated and normal myocardium and get an accurate estimate of chronic lesion dimensions. Shortening of T1 relaxation time of RF ablated myocardium observed in immediately post-ablation T1 Mapping and T1w scans was explained by temperature caused transformation of hemoglobin from ruptured and obstructed vessels into a paramagnetic agent¹. Our results indicate that this agent possibly has a relatively short lifetime or slowly washes out from the ablated regions or both