Re: Additional value of fetal magnetic resonance imaging in the prenatal diagnosis of central nervous system anomalies: a systematic review of the literature

G. Malinger,T. Lerman-Sagie

DOI: https://doi.org/10.1002/uog.14740

2015-02-01

Ultrasound in Obstetrics and Gynecology

Abstract:We read with interest the systematic review of the literature on the additional value of fetal magnetic resonance imaging (MRI) in the prenatal diagnosis of central nervous system (CNS) anomalies1. The review concludes that MRI supplements the information provided by ultrasound. The authors found that in 22.1% of cases MRI provided additional information, in 18.4% it disclosed CNS anomalies and in 30% the MRI findings were so different that the clinical management was changed. Since these results contradict our experience, described in our paper2 which was one of those included in this meta-analysis, we would like to review it critically and provide our insights into the results. Only two of the 13 studies included were performed after 2007, the year in which the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) published guidelines on sonographic examination of the fetal CNS. The authors also mention that only three studies described the ultrasound protocol and included orthogonal planes and the use of transvaginal probes. In contrast, the MRI sequences and the planes used were described in all cases. Their Table S1 describes the type of additional information provided by MRI1. The table shows clearly that the inaccuracy in the ultrasound diagnoses, in a considerable number of cases, was not the fault of the equipment but of the operator. In order to demonstrate our point, we analyzed the results of the first paper3. We chose this particular paper since it demonstrates the best ultrasound results, with a concordance of 94%. Even in this paper, it is difficult to explain the fact that ultrasound failed to diagnose anomalies that are readily demonstrable by ultrasound and have been reported in the literature, such as germinolysis4, irregularities of the ventricular wall5, a thin and short corpus callosum6,7, multiple subependymal heterotopias5, cavitations of the white matter8 and an occipital meningocele with inferior vermian agenesis9,10. In another two patients, one with white-matter cavitations and the other with increased white-matter signal on T2-weighted imaging, the MRI findings may be considered to be false positives, since the children were developing normally. This leaves only one of 185 cases in which MRI probably provided valuable additional information (‘presence of frontal sulcation abnormality’). The meta-analysis reported a 2.5% rate of false-positive and a 2% rate of false-negative interpretations. These percentages are extremely low and are in disagreement with some recent studies that specifically address this issue. Limperopoulos et al.11 found that, of 39 children with suspected posterior fossa anomalies who underwent prenatal and postnatal MRI, there was agreement in only 23 (59%)11. Postnatal MRI excluded the prenatal diagnosis in six (15%) and revealed additional findings in 10 (26%). The authors attributed these high false-positive and false-negative rates to the relatively early performance of the examinations. Moutard et al.12 described the follow-up of 17 children diagnosed in utero with apparently isolated agenesis of the corpus callosum and reported a false-positive rate of between 10% and 20%. Furthermore, Rossi and Prefumo1 give no additional information that could help explain some of the differences between the modalities13, such as operator experience and expertise in dedicated neurosonography and the number of cases diagnosed by ultrasound that did not require MRI. In conclusion, we think that the informed reader should understand that this systematic review ‘compares apples and oranges’; before we are able to conclude that MRI provides additional value to ultrasound, a large, multicenter study comparing contemporary dedicated neurosonography with MRI is required. In agreement with the conclusions reached by Salomon et al.3, we suggest that, at present, each center develops its own guidelines regarding the clinical indications for fetal CNS MRI, based on their personal experience and capabilities in the performance of ultrasound and MRI.