Perinatal survival and long-term neurodevelopmental outcomes of congenital chylothorax

BACKGROUND: Congenital chylothorax carries a poor prognosis, particularly with delivery at <32 weeks. Survival outcomes under consistent management prioritizing pregnancy prolongation despite worsening hydrops, as well as long-term neurodevelopmental data, remain limited. METHODS: In this single-center retrospective study, we analyzed fetuses with congenital chylothorax requiring perinatal drainage managed between 2005 and 2021. The protocol prioritized prolonging pregnancy to ≥32 weeks. Postnatally, octreotide and prednisolone were administered stepwise for persistent drainage. RESULTS: Of 65 fetuses (91% with hydrops), 29 (45%) underwent thoracoamniotic shunting. Fetal death occurred in 10 (15%) and infant death in 11 (17%). Among 55 live-born infants, 30 (55%) received octreotide and 25 (45%) received prednisolone. Of 28 survivors without major congenital anomalies (median gestational age, 35.9 weeks), 20 (71%) underwent Kyoto Scale of Psychological Development assessment at 3 years. Although the median Full-Scale Developmental Quotient (DQ) was 92 (interquartile range: 82-98), 40% had DQ < 85 (borderline or delayed). In the Language-Social domain, 55% had DQ < 85. The median Full-Scale Intelligence Quotient on the Wechsler Intelligence Scale for Children was 90 (interquartile range: 86-95). CONCLUSION: A protocol prioritizing pregnancy prolongation achieved favorable overall survival; however, long-term follow-up remains essential, particularly for Language-Social development. IMPACT: A perinatal management protocol for congenital chylothorax that prioritizes prolonging pregnancy beyond 32 weeks achieved overall mortality comparable to or lower than previous reports, despite a high prevalence of hydrops fetalis. Although the median Full-Scale Developmental Quotient at 3 years was within normal, 40% of survivors without major congenital anomalies showed borderline or delayed development, particularly in the Language-Social domain. A similar pattern was observed in school-age Intelligence Quotients. The median Full-Scale Developmental Quotient at 3 years did not differ between preterm and term infants, indicating that follow-up is necessary even for those born at term.