A Multimodal Score Accurately Classifies Fontan Failure and Late Mortality in Adult Fontan Patients

ObjectivesDespite the outstanding success of the Fontan operation, it is a palliative procedure and a substantial number of patients experience late failure of the Fontan circulation. Clinical presentation and hemodynamic phenotypes of Fontan failure are considerably variable. While various parameters have been identified as risk factors for late Fontan failure, a feasible score to classify Fontan failure and possibly allow timely risk stratification is lacking. Here, we explored the possibility of developing a score based on hemodynamic, clinical and laboratory parameters to classify Fontan failure and mortality. MethodsWe performed a retrospective study in our cohort of adult Fontan patients from two institutions [n = 198, median follow-up after Fontan 20.3 (IQR 15.6-24.3) years], identifying those patients with clinical Fontan failure (n = 52, 26.3%). Various hemodynamic, echocardiographic, laboratory and clinical data were recorded and differences between patients with and without Fontan failure were analyzed. We composed a Fontan Failure Score containing 15 parameters associated with Fontan failure and/or mortality and assessed its accuracy to discriminate between patients with and without late Fontan failure as well as late mortality and survival. ResultsLate failure occurred at a median of 18.2 (IQR 9.1-21.1) years after Fontan completion. Mortality associated with Fontan failure was substantial (25/52, 48.1%) with freedom of death/transplantation/take-down of 64% at 5 years and 36% at 10 years after onset of Fontan failure, respectively. Patients with Fontan failure had a significantly higher median Fontan Failure Score compared to non-failing Fontan patients [8 points (IQR 5-10) vs. 2 points (IQR 1-5), p < 0.001]. The score accurately classifies Fontan failure as well as mortality as assessed with receiver operating characteristic analysis. Area under the curve of the Fontan Failure Score was 0.963 (95% CI 0.921; 0.985, p < 0.001) to discriminate failure and 0.916 (95% CI 0.873; 0.959, p < 0.001) to classify mortality. ConclusionWe have developed an uncomplex yet remarkably accurate score to classify Fontan failure and late mortality in adult Fontan patients. Prospective validation and most likely refinement and calibration of the score in larger and preferably multi-institutional cohorts is required to assess its potential to predict the risk of Fontan failure and late mortality.