RÉSUMÉ
BACKGROUND: Temporary fetoscopic endoluminal tracheal occlusion (FETO) promotes lung growth and increases survival in selected fetuses with congenital diaphragmatic hernia (CDH). FETO is performed percutaneously by inserting into the trachea a balloon designed for vascular occlusion. However, reports on the potential postnatal side-effects of the balloon are scarce. This study aimed to evaluate the prevalence of tracheomalacia in infants with CDH managed with and without FETO and other consequences related to the use of the balloon. METHODS: In this multicentre, retrospective cohort study, we included infants who were live born with CDH, either with FETO or without, who were managed postnatally at four centres (UZ Leuven, Leuven, Belgium; Antoine Béclère, Clamart, France; BCNatal, Barcelona, Spain; and HCor-Heart Hospital, São Paulo, Brazil) between April 5, 2002, and June 2, 2021. We primarily assessed the prevalence of all (symptomatic and asymptomatic) tracheomalacia as reported in medical records among infants with and without FETO. Secondarily we assessed the prevalence of symptomatic tracheomalacia and its resolution as reported in medical records, and compared tracheal diameters as measured on postnatal x-rays. Crude and adjusted risk ratios (aRRs) and 95% CIs were calculated via modified Poisson regression models with robust error variances for potential association between FETO and tracheomalacia. Variables included in the adjusted model were the side of the hernia, observed-to-expected lung-to-head ratio, and gestational age at birth. Crude and adjusted mean differences and 95% CIs were calculated via linear regression models to assess the presence and magnitude of association between FETO and tracheal diameters. In infants who had undergone FETO we also assessed the localisation of balloon remnants on x-rays, and the methods used for reversal of occlusion and potential complications associated with balloon remnants as documented in clinical records. Finally we investigated whether the presence of balloon remnants was influenced by the interval between balloon removal and delivery. FINDINGS: 505 neonates were included in the study, of whom 287 had undergone FETO and 218 had not. Tracheomalacia was reported in 18 (6%) infants who had undergone FETO and in three (1%) who had not (aRR 6·17 [95% CI 1·83-20·75]; p=0·0030). Tracheomalacia was first reported in the FETO group at a median of 5·0 months (IQR 0·8-13·0). Symptomatic tracheomalacia was reported in 13 (5%) infants who had undergone FETO, which resolved in ten (77%) children by 55·0 months (IQR 14·0-83·0). On average, infants who had undergone FETO had a 31·3% wider trachea (with FETO tracheal diameter 7·43 mm [SD 1·24], without FETO tracheal diameter 5·10 mm [SD 0·84]; crude mean difference 2·32 [95% CI 2·11-2·54]; p<0·0001; adjusted mean difference 2·62 [95% CI 2·35-2·89]; p<0·0001). At birth, the metallic component was visible within the body in 75 (37%) of 205 infants with available thoraco-abdominal x-rays: it was located in the gastrointestinal tract in 60 (80%) and in the lung in 15 (20%). No side-effects were reported for any of the infants during follow-up. The metallic component was more likely to be in the lung than either outside the body or the gastrointestinal tract when the interval between occlusion reversal and birth was less than 24 h. INTERPRETATION: Although FETO was associated with an increased tracheal diameter and an increased probability of tracheomalacia, symptomatic tracheomalacia typically resolved over time. There is a higher risk of retention of metallic balloon components if reversal of the occlusion occurs less than 24 h before delivery. Finally, there were no reported side-effects of the metallic component of the balloon persisting in the body during follow-up. Longer-term follow-up is needed to ensure that no tracheal problems arise later in life. FUNDING: None.