Purpose or Case Report: Postmortem imaging is increasingly used following stillbirths and neonatal death. Estimating time of death has legal, biological, and ethical implications. This study aims to investigate feasibility of prenatal MRI in estimating time of death in intra-uterine fetal demise (IUFD).
Methods & Materials: Fetal MRIs performed between 2010 and 2019, containing the word “demise” in the report, and with at least one T2-weighted sequence including the brain of a demised fetus in the field of view were reviewed. Gestational age (GA) at demise was defined either as the day of procedures leading to IUFD or estimated GA of IUFD by ultrasonography. Based on the literature available on post mortem of stillbirths for estimating date of IUFD, the following parameters were adapted for MRI and evaluated: Opening of the mouth; Signal intensity of the globes; Appearance of the ventricles; Extra cranial soft tissue edema; Visualization of the cortical plate; Visualization of the laminar pattern of the fetal brain; Effacement of the sulci; Effacement of the pericerebral cerebrospinal fluid spaces; Collapse of the skull. Differences in time interval from fetal MRI to estimated IUFD (ΔT) for each imaging finding were explored with a Wilcoxon Sum of Ranks.
Results: In total, 44 fetuses met our inclusion criteria. Forty fetuses were monochorionic-diamniotic (MDA) twins, one dichorionic-diamniotic (DDA) twin, and three were singleton fetuses. Laser therapy was performed in 21 (48%) MDA fetuses and IUFD was spontaneous in remaining pregnancies. Median maternal age at time of MRI 30 years (IQR 27-35). Median GA at MRI was 24 weeks (IQR 23-28). Median GA at estimated IUFD was 20 weeks (IQR 18-22). Significant differences in ΔT (median in days; IQR) were identified for mouth opened (31;7) vs closed (25;7)(p=0.02); slit-like ventricles (34.5;18.25) vs non slit-like ventricles (25.5;10)(p=0.01); cortical plate visualization (27.5;13.25) vs non-visualization (39;13)(p=0.03); skull showing total collapse (37;15) vs mild/no collapse (23;7)(p=0.01). A trend association between ΔT and globes showing reduced signal intensity (34;10) vs preserved signal intensity (19;17) was found (p=0.05).
Conclusions: Opening of the mouth, non-visualization of the cortical plate, slit-like ventricles and total collapse of the skull are significantly associated with more days between IUFD and imaging. Estimating the time of IUFD is feasible by prenatal MRI.