OC05.03: Comparison of biochemistry analysis in fetal serum and fetal urine in the prediction of postnatal renal outcome in low urinary tract obstruction

Our aim was to compare the prognostic value of biochemistry in fetal serum and in fetal urine to predict renal outcome in a cohort of LUTO. We retrospectively studied the renal outcome following a prenatal diagnosis of LUTO in cases for which both fetal blood and fetal urine were sampled. Only cases in which the time interval between fetal blood and urine samples was less than 7 days were included. Fetuses that underwent vesicoamniotic shunting or fetal cystoscopy were excluded. Fetal urine was sampled from the bladder in all cases. We classified the renal outcome as either favourable, when postnatal renal function evaluated after a minimum follow-up of 18 months was normal (glomerular filtration rate upper than 59 mL/min/1.73 m2), or adverse in case of postnatal chronic renal failure or when renal histological lesions were present at autopsy in case of termination of pregnancy. Prognostic markers included serum β2-microglobulin in serum, and β2-microglobulin, sodium, calcium and chloride in urines. A prognostic model was constructed for urines and serum separately. These models were compared using ROC curves. Within the 50 cases included, the rate of adverse outcome was 34/50 (68%): autopsy confirmed severity of renal disease in all 27 cases of termination of pregnancy and 7/23 liveborn children developed renal failure. Fetal blood and urine sampling were performed at a median of 30 (ranges 19–35) weeks of gestation. In 45 cases, the two samples were performed at the same time. In the 5 others cases, the time interval was between 2 and 6 days. Whereas fetal serum and urine markers were all significantly associated with renal outcome (P<0.01), the ROC curves for the fetal serum and fetal urine prognostic models were similar (AUC=0.908 versus 0.904, P=0.91). Serum and urine biochemistry have a similar prognostic value in fetuses with LUTO.