Using MRI-derived observed-to-expected total fetal lung volume to predict lethality in fetal skeletal dysplasia.

Abstract

Pulmonary hypoplasia is the primary cause of perinatal death in lethal skeletal dysplasias. The antenatal ultrasound correlates for lethality are indirect, measuring the thorax (thoracic circumference, TC) or femur compared to the abdomen (TC/AC, FL/AC). A single study has correlated lethality with the observed-to-expected total lung volume (O/E-TFLV) on fetal MRI in 23 patients. Our aim was to define a cutoff value to predict lethality more specifically using MRI-derived O/E-TFLV. Two large fetal center databases were searched for fetuses with skeletal dysplasia and MRI; O/E-TFLV was calculated. Ultrasound measures were included when available. Each was evaluated as a continuous variable against lethality (stillbirth or death in the first month of life). Logistic regression and receiver operating characteristic (ROC) curve analyses evaluated the prediction ability. AUC, sensitivity, and specificity were calculated. P < 0.05 was considered statistically significant. A total of 80 fetuses met inclusion criteria. O/E-TFLV < 0.49 was a significant risk factor in predicting lethality, with sensitivity and specificity of 0.63 and 0.93, respectively, and an AUC of 0.81 (P < 0.001). FL/AC < 0.129 was also a strong variable with sensitivity, specificity, and AUC of 0.73, 0.88, and 0.78, respectively (P < 0.001). TC/AC and TC percentile were not significant risk factors for lethality. An O/E-TFLV of < 0.38 defines a specificity for lethality at 1.00. MRI-derived O/E-TFLV and US-derived FL/AC are significant predictors of lethality in fetuses with skeletal dysplasia. When prognosis is uncertain after ultrasound, calculation of MRI-derived O/E-TFLV may provide additional useful information for prognosis and delivery planning.