Prematurity and biliary atresia: a 30-year observational study

AIM OF STUDY: The diagnosis of biliary atresia (BA) remains challenging and delay can lead to significant morbidity with time to surgery a key factor in determining outcome. Prematurity may impact on outcome potentially delaying diagnosis. We sought to assess whether the premature BA infants (PBA) have a delayed time to surgery and as such, worse outcomes? METHODS: Review of a single-centre prospectively maintained database. Prematurity was defined as delivery < 37/40 gestation. PBA was compared with date-matched term biliary atresia controls on a 2:1 basis. Primary outcomes were clearance of jaundice (< 20 μmol/L) and native liver survival. A retrospective assessment of liver fibrosis was made on biopsies at diagnosis and at Kasai portoenterostomy (KPE) in both premature and term cohorts. Data are quoted as median (range) unless indicated. A P value of ≤ 0.05 was considered statistically significant. RESULTS: 21 (female n = 14, 67%) premature infants with BA were treated in the period Jan. 1988–Dec. 2016 and compared with 41 contemporaneous term BA controls. Median gestation was 33 (29–36) weeks and birth weight 1930 (948–4230)g. Twin pregnancy (n = 10) was the leading cause for prematurity and significantly higher than the controls (48 vs. 0%; P < 0.0001). Maternal co-morbidity was high (n = 10, 48%) including pre-eclampsia (19%) and diabetes (14%). Liver biopsy was performed in 19 (90%) patients (all diagnostic) at a median of 57 (4–266) days. Delayed diagnosis (> 50 days) was seen in n = 13 but not associated with parenteral nutrition use (46 vs. 33%, P = 0.59) or phototherapy (50 vs. 83%, P = 0.19). Both BASM (33 vs. 7.5%; P = 0.01) and duodenal atresia (19 vs. 0%; P = 0.01) were seen more frequently in the PBA cohort. Mean fibrosis scores (Ishak) from diagnostic biopsies were lower in the premature group than the control group (2.71 vs. 3.53, P = 0.043) indicating less fibrosis but this equalized by time of subsequent KPE (P = 0.17). Primary surgery was Kasai portoenterostomy (n = 20) at an older median age than controls (65 vs. 56 days; P = 0.06). Liver transplantation was the primary procedure in one late-presenting child. There was an increased but non-significant clearance of jaundice in the PBA group [n = 12/20 (60%) vs 20/41 (48%); P = 0.23] post-KPE. Native liver survival and true survival were not different (P = 0.58 and 0.23). CONCLUSIONS: PBA infants have similar outcomes to term infants, despite delayed diagnosis and higher frequency of the syndromic form. The high incidence of discordant twins supports the theory that epigenetic modifications could contribute to the pathogenesis of BA. LEVEL OF EVIDENCE: IIIc Retrospective Matched Cohort Study.