Variation in pulmonary vein size during the cardiac cycle: implications for non-electrocardiogram-gated imaging

Understanding pulmonary vein (PV) anatomy is important for the planning and execution of PV isolation for the treatment of atrial fibrillation, screening for PV stenosis after the procedure, and investigating the pathophysiology of atrial fibrillation. We hypothesized that significant changes in PV size occur during the cardiac cycle and sought to identify the relationship of data obtained with conventional non-electrocardiogram (ECG)-gated methods compared with ECG-gated measures of PV size using cardiovascular magnetic resonance.

A consecutive series of 14 patients in sinus rhythm were evaluated with non-ECG-gated contrast-enhanced magnetic resonance angiography and ECG-gated cine cardiovascular magnetic resonance of the PV. Pulmonary vein diameter, perimeter, and cross-sectional area (CSA) were measured using both methods.

Maximum diameter, perimeter, and CSA occurred simultaneously in all PV. The timing of the maximum size varied but generally occurred in ventricular diastole (101 +/- 112 milliseconds after mitral valve opening). The timing of minimum PV size also varied but generally occurred in ventricular systole (212 +/- 90 milliseconds before mitral valve opening). The difference between the maximum and minimum PV size was 15% +/- 8% for diameter, 15% +/- 7% for perimeter, and 27% +/- 12% for CSA (P < .001 for all). Contrast-enhanced magnetic resonance angiography correlated best with the ECG-gated maximum PV size (R2 > 0.48, P < .001 for all) and was greater than the minimum and average PV sizes (P < .05 for all).

All measures of PV size vary significantly during the cardiac cycle. Contrast-enhanced magnetic resonance angiography PV measurements correlate best with maximum PV size.