Potential clinical benefits and limitations of fetal virtopsy using high-field MRI at 7 Tesla versus stereomicroscopic autopsy to assess first trimester fetuses

Abstract

Objective: The aim of this study was to establish the diagnostic accuracy of high-field magnetic resonance imaging (MRI) at 7 Tesla (T) compared with that of stereomicroscopic autopsy for assessing first trimester fetuses. Methods: Nine consecutive cases of first trimester fetuses resulting from spontaneous and therapeutic pregnancy termination were considered. The cases were divided into two groups according to gestational age: the Embryo Group with cases of nine to 10 gestational weeks (GWs) and the Fetus Group with cases of 13 GWs. The first group was scanned using three-dimensional fast imaging with steady state precession (3D FISP), and the second group was scanned using a two-dimensional (2D) turbo spin-echo high-resolution T2-weighted imaging (T2 WI) protocol. A radiologist and two embryologists interpreted the images. All cases were evaluated by invasive autopsy, with pathologist blinded to the imaging results. In total, the database included 270 items for evaluation (9 cases × 30 structures/case). Results: The global agreement between fetal high-field virtopsy and microscopic or stereomicroscopic autopsy was evaluated using 225 evaluation items visible by both methods. Overall, using microscopic examination and stereomicroscopic autopsy as the gold standard, fetal high-field virtopsy had a sensitivity of 94.6% [95% CI, 87.2-98.3] and a specificity of 97.6% [95% CI, 95-98.8]. The positive predictive value (PPV) was 93% [95% CI, 85.7-96.6], and the negative predictive value (NPV) was 98.2% [95% CI, 95.7-99.4]. Cohen kappa coefficient of agreement was k = 0.92 [95% CI, 0.82-0.97], and the McNemar test showed p = 1.00. Conclusions: Virtual autopsy using high-field MRI at 7 T can be considered a safe alternative approach to stereomicroscopic autopsy for the assessment of fetal structural anomalies at the end of the first trimester of pregnancy.