Clinical and Molecular Diagnosis of Osteocraniostenosis in Fetuses and Newborns: Prenatal Ultrasound, Clinical, Radiological and Pathological Features

Osteocraniostenosis (OCS, OMIM #602361) is a severe, usually lethal condition characterized by gracile bones with thin diaphyses, a cloverleaf-shaped skull and splenic hypo/aplasia. The condition is caused by heterozygous mutations in the FAM111A gene and is allelic to the non-lethal, dominant disorder Kenny-Caffey syndrome (KCS, OMIM #127000). Here we report two new cases of OCS, including one with a detailed pathological examination. We review the main diagnostic signs of OCS both before and after birth based on our observations and on the literature. We then review the current knowledge on the mutational spectrum of FAM111A associated with either OCS or KCS, including three novel variants, both from one of the OCS fetuses described here, and from further cases diagnosed at our centers. This report refines the previous knowledge on OCS and expands the mutational spectrum that results in either OCS or KCS.


Introduction
A lethal disorder characterized by thin, fragile long bones and craniofacial dysmorphism was initially reported by Kozlowski and Kan in 1988 and Maroteaux et al. in the same year [1,2]. The name "Osteocraniostenosis" (OCS) was first suggested in 1994 by Verloes et al. [3] after the finding of a consistent set of clinical and radiological anomalies in two fetuses and one newborn and a re-evaluation of three cases described by Kozlowski and Kan and by Maroteaux in 1988 [1,2]. These anomalies included extremely thin bones; brachydactyly; dysmorphic facial features such as a small philtrum, short nose, narrow mouth with tented upper lip vermilion, dysmorphic ears and ocular malformations (microphthalmia, aniridia); short and overmodeled tubular bones of the hands and feet; cloverleaf shape of the skull; and spleen aplasia or hypoplasia. Both males and females have been reported.
In 2006 Elliott et al. [4] disputed the term OCS for the disorder by stating that craniosynostosis is not a common finding, even if the skull shape radiographically resembles the trefoil or cloverleaf skull ("Kleeblattschädel"). This distinguishing skull configuration would rather be caused by severe hypomineralization of the cranial bones. The term "osteocraniosplenic" syndrome instead of OCS was suggested.
While patients affected by OCS are usually stillborn or die shortly after birth, the allelic disorder known as Kenny-Caffey syndrome (KCS) shows a milder phenotype with impaired skeletal development, small and dense bones, short stature and no life-threatening complications, except for recurrent hypocalcemia with low levels of parathyroid hormone (PTH), which requires continuous therapy with calcium and activated vitamin D [5][6][7]. The facial phenotype of KCS is characterized by a triangular face with frontal bossing, delayed closure of fontanels, small eyes with hypermetropia and dental anomalies. Other features are brittle long bones with narrow diaphysis and radiographically noticeable medullary stenosis.
Kenny and Linarelli were the first to report this condition in 1966 [8], while in 1967, Caffey [9] described the same family (affected mother and son) in more detail, pointing to cortical thickening with medullary stenosis of long bones, osteosclerosis of cranial bones and sparing of pelvic and vertebral bones. Another family with dominant transmission of the Kenny syndrome, from an affected mother to one daughter and two sons, was reported by Majewski et al. in 1981 [10].
Subsequently, a superficially similar disorder was described that includes hypoparathyroidism, short stature and developmental delay, and is inherited as a recessive rather than a dominant trait [11][12][13][14][15][16]. While first described as "recessive variant of the Kenny-Caffey syndrome", when the gene responsible for this latter condition was isolated, it was named "Kenny-Caffey syndrome" (KCS); this is of course at odds with the original descriptions of the syndrome as a dominant trait. Currently, the OMIM catalogue perpetuates this error in listing "KCS type 1" (OMIM #244460) as recessive and linked to the TBCE gene, while the original, dominant KCS is listed as "KCS type 2" (OMIM #127000). In fact, the recessive KCS type 1 would better be called the Sanjad-Sakati syndrome, also known as hypoparathyroidism-retardation-dysmorphism syndrome (HRDS; OMIM #241410).
OCS and KCS share some common features: microphthalmia, a triangular face with frontal bossing, a spread of bone density associated with brittle and thin skeletal elements and hypocalcemia, which was not detected in Verloes's original report [3] but later noticed in most of the OCS patients who survived beyond the perinatal period [7]. This evidence has led Unger et al. [7] to believe that OCS and KCS might be allelic disorders, and this hypothesis was confirmed by the finding that all cases they studied had monoallelic de novo variants in FAM111A, which is located on chromosome 11q12.1. A previously unknown role of FAM111A in the regulation of PTH has been speculated, which would cause hypoparathyroidism and compromise normal skeletal development.
It has been argued by several authors [4,[17][18][19] that OCS is quite similar to Hallerman-Streiff syndrome (HSS). In fact, both of them show skull and long bone anomalies, but while HSS is not lethal, OCS often is; in addition, HSS has not been shown to have the characteristic facies of OCS or splenic hypoplasia.
Here we describe two clinical cases of fetuses with typical findings of OCS. Furthermore, we review the previous literature to define the most prominent prenatal ultrasound features, the main clinical and radiological aspects in the fetus and newborn and the most important features of the pathological examination.

Materials and Methods
All examinations and procedures described were conducted according to Good Clinical Practice and the Declaration of Helsinki (64th WMA General Assembly, Fortaleza, Brazil, October 2013). Data treatment was performed in accordance with Privacy Law (European regulation GDPR 2016/679). Radiological examinations, histopathological analysis and genetic testing were performed in the course of normal clinical practice and according to standard procedures, as described previously [7]. Signed informed consent was obtained from the subjects' parents for genetic testing and for the publication of pictures and clinical data.

Clinical Case #1
The first clinical case we considered was the fetus of a non-consanguineous couple (mother from Romania and father from Albania) whom we counseled during their first pregnancy. Prenatal ultrasound at 20 (+6) weeks of gestation found cloverleaf skull, microphthalmia, microretrognathia, thoracic hypoplasia, short long bones (<5th centile) and flexed feet. The karyotype on amniotic fluid was 46, XY.
Molecular analysis of the gene FGFR3 on amniotic fluid-requested because the ultrasound findings of micromelia and cloverleaf skull had led to the hypothesis of thanatophoric dysplasia type II-resulted normal. After termination of the pregnancy at 22 weeks, the fetus showed a weight of 320 g (<3rd centile), a crown-heel length of 24.5 cm (<3rd centile), a head circumference of 17 cm (<3rd centile), frontal bossing, a narrow forehead with supra-auricular bulging and parietal bossing, a wide anterior fontanelle (3 by 4 cm), short palpebral fissures, hypertelorism, a narrow mouth, low-set ears, camptodactyly (contracture of the proximal interphalangeal joints of the 2nd-5th fingers), a bilateral short hallux, a ventricular septal defect (VSD), micropenis and asplenia. The main features of the objective examination of the fetus are shown in Figure 1A.
Babygram X-rays displayed decreased skull ossification, 11 pairs of thin ribs, mild platyspondyly, mild cervical vertebral dysplasia, slender long bones with flared metaphyses and increased density due to stenosis of the medullary cavity ( Figure 1B).
Pathological examination of the femur revealed features similar to those reported in the literature [7], with grossly normal cartilaginous epiphysis but extremely thin diaphyses and broad metaphyses ( Figure 1C). Histologic examination of the growth plate of the proximal femur showed a normal epiphyseal resting cartilage with a rarefied proliferating zone and hyperplastic chondrocytes with narrow cartilage lacunae in the hypertrophic zone.
The combination of clinical and radiographic features suggested the diagnosis of OCS. Sanger sequencing of the FAM111A gene on DNA from fetal and parental blood revealed that the fetus had a de novo heterozygous variant NM_001312909.2:c.1026_1028del, NP_001299838.1:p.(Ser343del), already associated with OCS [7].

Clinical Case #2
The second case was the female fetus of a non-consanguineous couple. The pregnancy was terminated due to fetal anomalies. Labor was induced at the 21st (+2) week; the fetus was small for its gestational age, with a weight of 160 g (<3rd centile). The head was large; the cranial vault was undermineralized and most of its bones appeared to be absent (Figure 2A   The lateral ventricles of the brain were enlarged. The fetus featured small rudimentary ears, a flattened wide nose and prominent lips. Small effusions of edematous tissues were found in the pleural cavities and abdomen, along with edematous placenta. The body displayed short limbs, echogenic bowel and a small spleen. Whole-body X-rays ( Figure 2) revealed bilateral femoral fractures with bowing, as well as fractures with deformity of the left radius and ulna. Long bone length was reduced and was closer to 16 weeks of gestation. The ribs were gracile and irregular in appearance.
No definite callous formation was identified, but some of the irregularity may be due to fractures. The shafts of the long bones were thinned, with bowing of the fibula and ulna in particular. The metaphyses of the tibiae, femora and humeri appeared to be disproportionally enlarged.
A pathology report detailing the microscopic analysis of the brain indicated a maturation consistent with the stated gestation of approximately 21 weeks, but noted striking necrosis and calcification of the basal ganglia and of the thalamic region.
As in the previous case, evidence led us to consider the diagnosis of OCS as the most likely. Molecular testing via Sanger sequencing revealed a novel missense variant in FAM111A exon 5: NM_001312909.2:c.1542G>T, NP_001299838.1:p.(Met514Ile). We were unable to test the parents, but the same variant, which is absent in the reference population database gnomAD v2.1.1 (Genome Aggregation Database, gnomad.broadinstitute.org/, accessed on June 2021), had already been found de novo in another fetus with OCS during a diagnostic routine at the Lausanne laboratory (not available for publication). Therefore, according to the recommendations of the American College of Medical Genetics [20], the variant was classified as likely pathogenic.

Discussion
Both fetuses reported here had some typical features already associated with several other cases of OCS syndrome: intrauterine growth restriction (IUGR), dysmorphism of the head, face and ears with frontal bossing, bitemporal narrowing with supra-auricular bulging and parietal bossing, an undermineralized cloverleaf-shaped skull, thin long bones with increased density, flared metaphyses and obliteration of the medullary cavity, and a hypoplastic or absent spleen. The first case also had micropenis and camptodactyly.
We reviewed the previous literature to further delineate the main prenatal ultrasound, clinical, radiological and pathological features of OCS. We considered both the cases who had a molecular confirmation and those without analysis of the FAM111A gene but with a compatible clinical picture. Among the latter, several are actually familial cases, with similarly affected siblings from healthy parents and/or with established consanguinity of the parents. Although germinal mosaicism cannot be ruled out, it seems an unlikely explanation given the number of these occurrences. It was recently proposed that the pathogenic mechanism correlating FAM111A defects with OCS involves PTH dysfunction or deficiency [7]. This may lead to the reasonable conjecture that defects in other genes involved in embryonic development and with a prominent role in PTH synthesis/regulation can engender a similar phenotype, possibly with an autosomal recessive mode of inheritance. Table 1 lists the clinical characteristics reported in OCS and their occurrence out of the number of cases evaluated for the specific feature, counting separately for FAM111Apositive cases, all non-tested cases and non-tested cases without suspicion of recessive inheritance. The types of examination that may be employed to detect each feature are also indicated.
Based on the data from cases with FAM111A testing, fragile bones with defects in skull shape and ossification, along with growth delay, are the most distinctive characteristics of OCS. Defining the other notable features based on the clinical reports is not straightforward: some cases are newborns, while most are fetuses at different times of gestation, with some showing different degrees of maceration. Therefore, we choose to exclude a few non-tested cases which had insufficient data and did not have a definite brittle bone aspect. presented in bold when at least 50% of the cases were evaluated; d presented in bold when at least 9 cases were evaluated; e i.e., excluding siblings and other cases with suspected recessive inheritance; f detection possible, but not always optimal; g usually with inverted U/V-shaped mouth; h at birth; at least 8 known to be intrauterine; i irregular shape/distribution of chondrocytes; j depending on the type of brain anomaly; † deceased at given age.
Head and face characteristics were quite consistent, with most fetuses and newborns showing a combination of frontal bossing, supra-auricular bulging, a flat face, short nose and low-set ears. Microphthalmia is a less common clinical feature; however, it can be helpful to suspect the diagnosis.
Micropenis in males was a consistent finding among the few cases with molecular confirmation of a FAM11A variant [7], including our clinical case #1. In previous reports it had been clearly documented in only a single case [4]; however, data were only available for less than half of the male fetuses/newborns considered. This suggests that micropenis might be a significant but underreported feature of OCS.
The radiological evidence was the most evocative aspect of OCS. In addition to an often-dysmorphic appearance, the skull usually showed markedly delayed ossification, frequently with missing bones and/or a wide anterior fontanelle. In nearly all reported cases, the long bones had a distinctive aspect with diaphyseal stenosis and metaphyseal flaring. All but one reports indicated thin ribs, and platyspondyly was observed in at least two-thirds of the cases. Bone fractures at birth, a consequence of these ossification defects, occurred less frequently but can be very useful for the diagnostic process.
Pathological examination can reveal important anatomical and histological aspects. Splenic hypoplasia or aplasia was the most common finding after the skeletal and craniofacial features and it may be useful to suspect this condition. Defects in other organs were present but less frequent, and included lung hypoplasia, hepatomegaly or other liver dysfunctions and congenital heart disease. Thymic hypoplasia was also present but had only been looked for in a small number of cases, and further investigations are required to establish its frequency. Similarly, extramedullary hemopoiesis might also be a poorly looked-for but consistent feature of OCS, since it was observed in 2 out of 6 cases with confirmed FAM111A mutation and 2 out of 4 cases with no molecular testing.
Along with the features revealed by prenatal ultrasound and babygram X-rays, the histological examination of bone growth plates is emerging as a very important criterion to support the diagnosis of OCS, as Spear et al. reviewed [31]. Several of the examined cases showed varying degrees of growth plate irregularity, generally involving short/disorganized columns of pleomorphic chondrocytes in the proliferating and/or hypertrophic zone. Macroscopic and histological examination of a femur from our clinical case #1 was well within the spectrum of abnormalities described so far. In fact, it showed marked similarities with case #7 from Unger et al. [7]: both displayed a thin diaphysis and grossly enlarged metaphyses, and the histological aspect of the growth plate was mildly disorganized ( Figure 1C,D).
KCS patients, recently reviewed by Cheng et al., have a milder phenotype overall [32]. They share elements of defective mineralization with OCS, but at the milder end of the spectrum, with features such as delayed closure of the anterior fontanelle rather than severely decreased skull ossification [7,32]. KCS patients are characterized by hypoparathyroidism and hypocalcemia; low calcium levels have also been detected in OCS cases, supporting the hypothesis of a PTH dysfunction underlying both conditions [7]. Figure 3 lists the OCS-and KCS-associated FAM111A variants reported in the literature, and novel variants observed during diagnostic activity at the Lausanne Genetics Laboratory. These include the Met514Ile variant also described here in clinical case #2, as well as the de novo variants Phe520Cys, seen in a child with KCS not reported before, and Phe311Ile, identified in an OCS case with IUGR, macrocephaly and facial dysmorphisms (ClinVar SCV001439356.1). Most variants are clustered together near the protein C-terminus, and a definite genotype-phenotype correlation, explaining the different severity of OCS and KCS, has not been established yet.  , bottom). Cys485Phe was reported in [33], Ser541Tyr/Pro were reported in [5,32], respectively, Tyr562Ser was reported in [21], while all other variants were observed at the Laboratory of Genetics at CHUV ( [7] and this report). The domain in orange has homology to trypsin-like peptidases, including an untested catalytic triad purportedly composed of Ser541/His385/Asp439 (red bars).
Through homology comparison and functional studies, the protein encoded by FAM111A is emerging as a serine protease putatively involved in DNA replication and response to replication stress [34][35][36][37][38]. FAM111A colocalizes with the replication fork, possibly interacting with single-strand DNA through its central region, while its C-terminal portion features a conserved Ser/His/Asp catalytic triad and acts as a chymotrypsin-like peptidase ( Figure 3). Although its exact biological targets are still being investigated, auto-cleavage activity has been documented, both in cis and in trans, suggesting that multiple copies of FAM111A might interact with other factors to form active complexes [36][37][38]. Interestingly, the SWISS-MODEL server reports a high homology with serine protease HtrA2, which is known to function as a trimer [39].
Haploinsufficiency is unlikely to explain the pathogenetic mechanism. As already noted [7], null alleles are present in reference population databases, including the latest releases of gnomAD (v2.1.1, v3.1.2, accessed June 2021), and have not been associated with known cases of OCS or KCS. In vitro experiments on FAM111A knockout cell lines indicate a negligible effect on cell viability, except for an increased sensitivity to specific agents inducing DNA replication stress [36,37]. Future studies on model organisms (currently in planning, according to the International Mouse Phenotyping Consortium) are needed to support these observations.
All OCS-or KCS-associated variants identified to date lead to single amino acid substitutions or deletions. Based on 3D models predicting that most variants occupy a limited area on the protein surface in the vicinity of the protease domain, a dominant gain-offunction pathogenetic mechanism was proposed [7]. Accordingly, recent experimental studies suggested that known disease-associated variants increase an intra-and intermolecular autocleavage mechanism and display protease activity-dependent cytotoxicity [37,38]. It is tempting to speculate that OCS-associated variants occurring near the cleavage site might influence autoproteolytic specificity (e.g., Ser342), while variants clustered in the C-terminal peptidase-homology domain might affect its catalytic activity [37]. Despite these advances, functional evidence on FAM111A is still limited, in particular concerning its targets and the subcellular localization of its variants, and has not yet provided an explanation for the gap in severity between OCS and KCS. Homology-and prediction-based structural models, such as those generated by AlphaFold (accessed July 2021) [40], show high confidence only in a restricted area that does not encompass all reported variants, and the putative catalytic triad has not been directly tested yet [38]. Further investigations will be needed, as well as the description of a larger number of cases and their associated variants, in order to define the pathogenetic mechanism of these allelic conditions.

Conclusions
OCS can be recognized prenatally in the presence of evocative elements such as craniostenosis or cloverleaf skull; thin and brittle bones; asplenia, usually co-occurring with the nonspecific findings of IUGR; and limb undergrowth.
Molecular confirmation of the diagnosis is essential for the parents as it allows geneticists and other healthcare professionals to establish a low recurrence risk, and to plan or limit prenatal investigations accordingly. Institutional Review Board Statement: Ethical review and approval were waived for this study because, according to the policy of "AUSL-IRCCS di Reggio Emilia" and "Lausanne University Hospital", informed consent from the patient (or legal caretaker) is considered sufficient for Case Reports of an observational nature.

Informed Consent Statement:
Written informed consent to publish this paper was obtained from the parents of all subjects involved in the study.

Data Availability Statement:
All data is available upon reasonable request by contacting the corresponding author.