Clinically significant posture-related oxygen saturation variability in Fontan circulation: prevalence and determinants.

BACKGROUND

Patients with Fontan circulation often have low baseline oxygen saturation (SpO2), and posture-related SpO2 variability is sometimes observed in clinical practice. However, the prevalence and determinants of clinically significant posture-related SpO2 variability have not been systematically investigated.

METHODS

This observational study included 103 patients with Fontan circulation and 166 patients after biventricular repair. After 1:1 propensity score matching by age and sex, 103 patients in each group were analyzed. SpO2 was measured in the sitting and supine positions, and the difference between supine and sitting SpO2 (ΔSpO2) was calculated. Clinically significant posture-related SpO2 variability was defined as an absolute ΔSpO2 of ≥3%. Clinical, hemodynamic, and imaging variables were compared between Fontan patients with and without significant variability. Multivariable logistic regression analysis was performed to identify independent determinants.

RESULTS

All patients in the biventricular group showed SpO2 variability of ≤2%. In contrast, 28 of 103 Fontan patients (27%) had clinically significant posture-related SpO2 variability. SpO2 increased in the supine position in 26 patients and decreased in 2 patients. Patients with significant variability were older and had a longer duration since the Fontan procedure. Veno-venous collaterals (VVCs) were more prevalent in these patients, whereas pulmonary arteriovenous malformations (PAVMs) were not more prevalent. Five patients with significant variability had neither VVCs nor PAVMs. In multivariable analysis, VVCs were independently associated with clinically significant posture-related SpO2 variability (odds ratio 4.85, 95% confidence interval 1.74-13.50, p = 0.003).

CONCLUSIONS

Clinically significant posture-related SpO2 variability occurs in approximately one-quarter of patients with Fontan circulation but not in patients after biventricular repair. This variability is independently associated with VVCs, although clear hemodynamic compromise was not demonstrated and mechanisms other than detectable right-to-left shunts may also contribute.