Introduction

There is a growing number of publications that describe the range of appearances of cortical formation anomalies (CFA) in fetuses on in utero and post mortem magnetic resonance (MR) imaging [1,2,3,4,5] and it is well known that CFA are frequently associated with other brain malformations, including complete agenesis of corpus callosum (ACC). Conversely, when complete ACC is detected on prenatal imaging a detailed search should be made for CFA because of the high association and substantial effect on prenatal counselling. In a previous publication, we designed and applied an MR-based classification system for CFA based on laterality, symmetry, and fine detail of the CFA in a cohort of over 350 fetuses [6]. During that study, we noticed a cluster of fetuses with unilateral CFA and ACC.

We have focused on fetuses with complete ACC and unilateral CFA in this study in order to provide a detailed morphological analysis to define the imaging characteristics of such fetuses. CFA are well known to be possibly associated also with partial agenesis or CC hypogenesis; however, the definition of complete ACC is relatively straightforward and not influenced by possible ambiguity related to the high variability of residual segments. The primary aim of the study is to determine if there are indicators (such as the gender of the fetus) that could be used to delineate a genetic substrate of the phenotype in order to inform future studies.

Materials and methods

The ethical approach and full methodological details for locating and analysing the original cohort have been described in detail previously [6] but are summarised here. The MR imaging data of fetuses with CFA were collected over an 18-year period (2000–2017) from seven centres in Italy and England, which included approximately 11,000 MR examinations of the fetal brain. All of the studies were performed on 1.5-T scanners and all fetuses had T2-weighted (single-shot FSE) images in the three orthogonal planes (3–5 mm thickness) as a minimum requirement. An expert consensus panel (ECP) was formed to confirm and classify the type of CFA present using the flow diagram developed by the authors in which divided the cases into two main classes based on bilateral or unilateral involvement [6]. The ECP were also asked to record other associated brain abnormalities, including ACC. There were 43 fetuses with ACC and unilateral CFA in the original cohort, and along with two new cases, our current report consists of 45 fetuses with ACC and unilateral CFA. For the present descriptive work, two experts in fetal MR imaging (A.R., C.P.) re-evaluated those 45 studies and provide a more nuanced anatomical description based on lobar location and laterality of the CFA. Other associated anomalies were evaluated including sex of the fetus as from ultrasound, intrauterine MR imaging, fetal karyotyping, autopsy, and post-natal data (at least one criterium available for all cases but one, which was lost at follow-up after unclear pelvic view by ultrasound and MR imaging); head size by biparietal and anterior-posterior diameter measurement according to ref. [7, 8]; and the presence of interhemispheric cysts, ventriculomegaly, and posterior fossa malformations.

Statistical analysis

All analyses were carried out using Statistical Package for Social Science (IBM SPSS Statistics 25). Differences between males and females were probed for categorical and numerical differences. Distributions of discrete and continuous variables were examined for normality using the Shapiro-Wilk test and inspection of histograms. Group differences were tested using either unpaired Student t-test or Mann-Whitney U-test depending if the data appeared to be normally distributed or not. Associations between dichotomous variables and categorical with a level five were studied with contingency tables and tested using the Chi-square test or, alternatively, using the Fisher’s Exact test depending on the frequencies of the contingency tables. When statistical significance was found in contingency tables greater than 2 × 2, pair-wise post hoc tests were performed. Correlation between discrete and numerical variables was studied through the Spearman Correlation Coefficient while dependence between categorical variables was assessed with the Chi-square test. Within-group differences between percentages were assessed by the chi-square test or Fisher’s Exact test, according to the frequencies of the contingency tables. Two-tailed p value < 0.05 was considered statistically significant. p value for post hoc test was adjusted with Bonferroni correction.

Results

Table 1 summarises the main demographic data and the fetal MR imaging findings of the 45 fetuses. The median gestational age (GA) at the time of the MR exam was 22 weeks, interquartile range = 4.25 weeks, Q1 = 21 weeks, Q3 = 25.25 weeks, and full range (20 to 36 weeks).

Table 1 Demographic data and the fetal MR imaging findings of the 45 fetuses cohort
Full size table

CFA laterality and lobar involvement

The CFA involved the frontal lobe in 42/45 fetuses (93%) and was confined to the frontal lobe in 18 of those fetuses. The parietal lobe was involved in 25/45 fetuses (56%) including two cases where the CFA involved the parietal lobe only. The temporal lobe was involved much less frequently, in only 3/45 fetuses (7%), and in all of those cases, other lobes were involved. The occipital lobe was involved in only 1/45 fetus (2%), and in that case, the whole hemisphere was involved. The frontal lobe was statistically significantly more frequently involved than other lobes (p < 0.001).

CFA types and fine morphology.

The CFA types according to the scheme reported in ref. 6 were in 17/45 cases C1-type, in 13/45 C2, and in 15/45 C5. The detailed morphology encompassed at least one of the following patterns: multiple tiny excessive invaginations of cortical surface; few (two or three) abnormally major invaginating sulci; irregular fine profile “wart-like”; large tangle of multiple chaotic and deeply invaginating sulci (Figs. 1, 2 and 3). It has to be taken into account that the fine appearance of CFA may be also influenced by the GA [4], since for example, few focally invaginating sulci in a still “lissencephalic” fetus at 21 weeks of age may evolve towards a large tangle of abnormal gyri in a highly gyrificated late third-trimester brain.

Fig. 1
figure 1

A prenatal ss-FSE T2-weighted image from nine exemplificative male cases with CFA located in the right hemisphere is respectively reported (GA in weeks is inserted). Arrows point the CFA location. The frontal lobe alone or in association with parietal lobe is almost always involved. CFA can appear as one of the following: multiple tiny excessive invaginations of cortical surface; few (two or three) abnormally major invaginating sulci (20 and 25 weeks cases example); irregular fine profile “wart-like”; large tangle of multiple chaotic and deeply invaginating sulci

Full size image
Fig. 2
figure 2

A prenatal ss-FSE T2-weighted image from nine exemplificative male cases with CFA located in the left hemisphere is respectively reported (GA in weeks is inserted). Arrows point to the CFA location. The frontal lobe alone or in association with the parietal lobe is almost always involved. CFA can appear as one of the following: multiple tiny excessive invaginations of cortical surface (23 weeks case example); few (two or three) abnormally major invaginating sulci; irregular fine profile “wart-like”; large tangle of multiple chaotic and deeply invaginating sulci (34 weeks case example)

Full size image
Fig. 3
figure 3

A ss-FSE T2-wighted prenatal image example from each of the seven male cases with detection of an interhemispheric cyst is respectively reported (GA in weeks is inserted). Arrows point to the unilateral CFA. Asterisks indicate the interhemispheric cyst (in the third case the cyst is not visible on the same section of the CFA). In the last (seventh) case an insert better depicts the CFA

Full size image

Gender of the fetus

The sex of the fetus was known in 44/45 cases with 36 male and 8 female fetuses indicating a statistically significant excess of male fetuses with ACC and unilateral CFA (male:female = 4.5:1; p < 0.0001—chi-squared test). There were no statistically significant differences between male and female fetuses in terms of GA at MR (p = 0.438) or head size (p = 0.289).

There was no statistical difference between the laterality of the CFA in male fetuses (right hemisphere in 22/36 (61%), left hemisphere in 14/36 (39%)—p = 0.18). CFA lobar location within the hemisphere was in the order of frequency: frontal-parietal (16/36 cases), frontal (16/36 cases), parietal (2/36 cases), temporal (1/36 case), frontal-temporal (1/36 case), and frontal-parieto-temporo-occipital (0/36 cases). The frontal lobe (either in isolation or along with other lobes) was significantly more frequently involved when compared to other locations (p < 0.001). CFA was in the mesial aspect of the hemisphere in 17/36, in the lateral in 2/36, and involved both aspects in 17/36 cases. By pooling together isolated mesial and mesial-lateral extension, they resulted in significantly prevalent (p < 0.001) (Figs. 1, 2 and 3).

Statistical analysis on CFA laterality and lobar involvement in the eight female fetuses could not be performed because of the small number of cases.

Associated intracranial abnormalities

Ventriculomegaly was present in 16/45 fetuses, 15 males and one female (not a statistically significant feature—p = 0.125). Interhemispheric cysts were present in 12/45 cases: 7 males, 4 females, one sex unknown, hence interhemispheric cysts were present in half of the female fetuses but only 16% of male fetuses but this cannot be considered statistically valid because of the small numbers of female fetuses (Fig. 1 supplementary online material). None of the fetal cases in the cohort had clastic lesions using our previously published criteria [6].

Discussion

We have reported a cohort of fetuses with unilateral CFA and ACC, which is very likely to represent a condition (or a small group of conditions) caused by genetic anomalies, because of the absence of any clear imaging signs of acquired pathology and the very high prevalence of males (80%). In those fetuses (particularly in the male fetuses) the unilateral CFA was preferentially located in the frontal lobes, either in isolation or extending into the parietal lobe in its mesial aspect. This quite characteristic and repetitive location of the CFA does not favour the hypothesis of a clastic aetiology which is usually related to more randomly located anomalies; it rather supports the genetic aetiology hypothesis. We cannot currently provide insights into genetic substrate since many genetic anomalies may interact and cause CFA; the striking male gender preponderance might suggest some X-linked entity, at least for part of our cases.

We believe that the identified cohort carries a very low possibility of selection bias: our cases come originally from a MR database of about 11,000 fetal MR examinations gathered from seven international centres [6]. Indeed, the present study population represents the final result of a truly random case collection process, that was prompted by the current trend to perform prenatal ultrasound screening in developed countries at around 20-week gestation. Such screening is usually followed in ACC-suspected cases by an expert ultrasound exam, which in our population, albeit in a minority of cases, identified some unilateral cortical irregularities indeed. Our data may encourage prenatal sonographers to strengthen their scrutiny for associated CFA, especially in male fetuses with ACC. In this regard, the 3D-ultrasound approach may overcome the limitations of the uneven echogenicity between the two hemispheres affecting 2D cross-sectioning. In the vast majority of our population the identification of the CFA was the result of the diagnostic process triggered by ultrasound; therefore, in cases without the ACC “red flag” CFA may go undetected in-utero. Meanwhile, we cannot totally rule out that additional smaller CFA in other locations might have been overlooked by in-utero MR imaging since post-natal imaging or autoptic gold standard was not available in many cases.

While interhemispheric cysts and ventriculomegaly are frequent, findings in ACC and their common pathophysiological basis are quite straightforward; focal CFA are not as frequent and easily explicable as well. It is not clear whether the ACC and the unilateral CFA are associated because of a common pathophysiology mechanism or if the CFA is causally related to the ACC for example by interfering with the normal axonal guidance that is required to form the normal corpus callosum.

The present study concentrates on “complete” ACC and we deliberately did not include cases in which the corpus callosum was formed in part (i.e. “hypogenesis of the corpus callosum”). The rationale for this was that we wanted to focus our analysis on clear-cut, unambiguously defined cases, which is possible for complete ACC because the definition of the entity is relatively straightforward. In contrast, there can be substantial ambiguity about definitions of cases in which some, but not all, of the corpus callosum, is present (hypogenesis of the corpus callosum, dysgenesis of the corpus callosum, callosal thinning, etc.) and this problem has been discussed in the recently published literature [9]. We can report, however, in the cohort of > 350 fetuses with CFA there were 16 fetuses with unilateral CFA and hypogenesis of the corpus callosum, as defined by the original ECP. The gender of those fetuses was known in 11/16 (69%) and a male excess was found (8 males and 3 females). The male preponderance of approximately 2.7:1 in that group is not as high as in the cases of unilateral CFA and ACC (4.5:1) but it still indicates a substantial excess of male fetuses.

Our observation regarding the excess male prevalence is consistent with the small volume of existing publications of case reports/small case series [10,11,12] of unilateral CFA. Meanwhile, in other larger population studies on ACC with or without associated anomalies males result only mildly to moderately more frequently affected than females [13,14,15]. However, it has to be underlined that in the latter larger population studies ACC is frequently associated with any sort of intracranial or somatic anomalies, so that data is quite unspecific. Instead, if we consider the small case series studies with specific association with CFA [10,11,12], such reported cases share their core features more closely with the ones of our cohort. Albeit in those reports, interhemispheric cysts were frequent findings, whilst in our population, interhemispheric cysts were detected in only 12/45 cases overall and proportionally less in male cases (7/36). We may have missed the detection of some interhemispheric cysts, since we relied on fetal MR imaging only and this should be taken into account; still, the large part and at least the majority of our male cohort did not have interhemispheric cysts. Therefore, the hypothesis that an interhemispheric cyst may be involved in halting the normal corpus callosum formation is far from being an exhaustive explanation at least for the majority of our male cases. Ventriculomegaly was present in almost half of the males but it is well-known that some degree of ventriculomegaly is a common feature in cases of isolated ACC.

No significant correlation between morphological features and different gender was found; however, statistical analysis was difficult because of the small number of female fetuses, Regarding the resulting small group of eight female fetuses, it should be mentioned that the possibility that some cases carried Aicardi syndrome is likely, since for example half of them showed clear interhemispheric cysts (Fig. 1 supplementary online material). Due to the small number of females, the comparison with males has low statistical power, even if, as just stated, the proportion of cases with interhemispheric cysts was definitely higher in the females.

There are several limitations to the current study including the retrospective nature and the lack of genetic, histological, and follow-up clinical data. In addition, we do not possess enough material to pursue any statistically valid clinical follow-up assessment on such cohort (mainly due to the high rate of terminations of the pregnancy); however, we provide as an appendix to our report (appendix-supplementary online material) some of the follow-up stories we were able to collect, since we believe that the anectodical cases with the discrepancy between the putative severity of the prenatal imaging picture and the not as severe clinical outcome deserve to be mentioned.

Conclusions

Our report highlights the common MR morphological features in such cases and we propose that such fetuses are very likely to have genetic anomalies and further laboratory-based research is required to define the anomalies and investigate their clinical significance.

Based on our data, when ACC is detected at prenatal imaging in a male fetus, unilateral CFA in the frontal and parietal lobes might be a possible specific additional finding.

Despite the rarity of this “not isolated ACC” condition, the relatively high fraction of cases undergone pregnancy termination in our historical series urges the need for further investigation on pathology or conversely outcome correlates, in order to provide a full rationale counselling.