Improvement in long-term survival after hospital discharge but not in freedom from reoperation after the change from atrial to arterial switch for transposition of the great arteries

Objective: To compare survival, freedom from reoperation, and functional status between atrial switch and arterial switch operations for transposition of the great arteries. Methods: Data from 88, 329, and 512 patients who underwent Mustard, Senning, and arterial switch operations between 1974 and 2006 were analyzed. Results: In-hospital mortalities were 8.0% for Mustard, 4.6% for Senning, and 6.4% for arterial switch. Presence of ventricular septal defect (hazard ratio 3.3, P < .001) was the only risk factor for in-hospital mortality in multivariate analysis. Follow-up for Mustard was 22.6 ± 8.1 years, for Senning was 18.2 ± 5.7 years, and for arterial switch was 9.5 ± 5.7 years. Highest survival at 20 years was after arterial switch (96.6% ± 1.3%), followed by Senning (92.6% ± 1.5%) and Mustard (82.4% ± 4.3%). Transposition with ventricular septal defect (hazard ratio 3.1, P < .001), transposition with ventricular septal defect and left ventricular outflow tract obstruction (hazard ratio 3.0, P = .029), and Mustard operation (hazard ratio 2.1, P = .011) emerged as risk factors for late death, with arterial switch a protective factor (hazard ratio 0.3, P = .010). Highest freedom from reoperation at 20 years was after Senning (88.7% ± 1.9%), followed by arterial switch (75.0% ± 6.4%) and Mustard (70.6% ± 5.4%). Presence of complex transposition (hazard ratio 2.1, P < .001), previous palliative operation (hazard ratio 1.8, P = .016), surgery between 1985 and 1995 (hazard ratio 2.6, P = .002), surgery after 1995 (hazard ratio 3.5, P < .001), and Mustard operation (hazard ratio 3.3, P < .001) emerged as risk factors for reoperation. Conclusion: Change from atrial to arterial switch led to improved long-term survival after hospital discharge but not to lower incidence of reoperation. Survival and freedom from reoperation are determined by morphology.