Authors:
G. Bastarrika, J. Arias, M. Ferreira, M. Arraiza, G. Viteri, G. Rabago; Pamplona/ES
DOI:
10.1594/ecr2010/C-0548
Results
Patient population
Twenty-nine consecutive heart transplant recipients (27 men, 2 women) with a mean age of 64.1±13 years (age range 26–80) underwent cardiac DSCT and MRI. Patient characteristics are shown in Table 1. Overall mean time from transplantation to study enrolment was 122.8±69.7 months (range: 37–236 months). The mean heart rate was 85.9 ± 10.1 beats per minute (bpm) during DSCT (range: 68-107) and 86.1 ± 9.5 bpm (range: 64-114) during MRI (p=0.91, N.S). Among the twenty-nine orthotopic heart transplant recipients thirteen were operated with the standard biatrial technique and sixteen were transplanted following the bicaval technique. Mean time from transplantation to study enrollment significantly differed between the two groups, being 190.7±38.2 months (range: 123–236 months) for individuals operated with the standard biatrial technique and 67.6±25.6 months (range: 37–114 months) for patients with bicaval anastomosis (p<0.001). No significant difference was observed in age distribution, body mass index (BMI) or heart rate during data acquisition.
Left atrial volumes and function
Cardiac DSCT and MRI exams were successfully accomplished in all patients without complications. Left atrial ED was determined at 35% of the RR interval in 8 subjects (27.6%), 40% of the RR interval in 18 (62.1%) and 45% of the RR interval in 3 individuals (10.3%). Left atrial ES was selected at 95% of the cardiac cycle in all individuals. All studies were suitable for analysis and allowed adequate manual segmentation for volumetric assessment. The mean time required for manual contour tracing in DSCT (462.8±69 s) and MRI (402.5±78.4 s) exams was similar (p=0.18).
Mean values for left atrial volumes and ejection fraction are summarized in Table 2. A systematic overestimation of left atrial volumes by DSCT when compared with MRI was observed. Left atrial volumetric quantification as determined with DSCT revealed a mean EDV of 170.9±78.1 ml and a mean ESV of 139.5±76.6 ml, whereas MRI showed a mean EDV of 158.2±72.5 ml and a mean ESV of 124.2±68.2 ml (p= 0.012 for EDV and p<0.01 for ESV). The mean EF estimated on DSCT and MRI was 20.8±7.5% and 23.6±7.7%, respectively (p<0.01). There was excellent correlation between DSCT and MRI measurements for all left atrial parameters (r ≥0.88). Bland-Altman analysis demonstrated a trend toward DSCT resulting in slightly lower values for left atrial ejection fraction (mean difference 2.8 ± 3.8%) and higher EDV (mean difference 12.7 ± 25.5 ml) and ESV (mean difference 15.2 ± 22.8 ml) with respect to MRI. Bland and Altman plots and CCC indicated excellent agreement between DSCT and MRI left atrial volumes and function measurements (CCC≥ 0.82).
Separate analysis of the two transplantation surgical techniques showed statistically not significant differences in EDV and ESV quantification with DSCT with respect to MRI in individuals operated with the standard biatrial technique. In patients with bicaval anastomosis overestimation of left atrial volumes and underestimation of the ejection fraction was observed with DSCT compared to MRI.
Comparison of transplantation techniques
Differences in left atrial volumes and function according to the cardiac transplantation technique are shown in Table 3. Both DSCT and MRI demonstrated significantly larger EDV (212.1±97.6 ml and 200.1±88.6 ml) and ESV (181.7±94.4 ml and 164.6±82.9 ml) in individuals who underwent heart transplantation with the standard biatrial technique as compared to patients with bicaval anastomosis (EDV: 137.5±33.4 ml and 124.2±27.9 ml; ESV: 105.1±32.1 ml and 91.5±24.9 ml) (p<0.05). Left atrial ejection fraction turned out to be significantly higher in the latter group as compared with individuals operated following the standard biatrial surgical technique (mean difference of 8.4% and 7.2% estimated with DSCT and MRI, respectively) (p<0.01).
Interobserver agreement
The interobserver agreement for left atrial parameters was excellent by means of DSCT (0.94 for EF, 0.99 for EDV, 0.99 for ESV) and using MRI (0.80 for EF, 0.96 for EDV, 0.95 for ESV) (Table 4). The difference in interobserver agreement between DSCT and MRI was not statistically significant (p>0.05).