Editor—Craniolacunia (“Lückenschädel”) describes rounded or finger-like defects of the inner table of the membranous portion of the skull that are surrounded by strips of normal bone.1 In most cases, craniolacunia is associated with spina bifida and a resulting myelo- or myelomeningocele. Additional associated anomalies include hydrocephalus, Arnold-Chiari malformation (congenital hindbrain hernia), and Klippel-Feil sequence (fusion of cervical vertebrae/hemivertebrae).2 The pathological mechanism resulting in craniolacunia is not known. Since depressions in the inner table of the calvarium are particularly deep in cases of oxycephaly, scaphocephaly, and brain tumour, increased intracranial pressure was invoked as an aetiological factor.1 Increased intracranial pressure, however, can give rise to impressiones digitatae (impressions of cerebral gyri) but does not in itself result in craniolacunia. Thus, craniolacunia has not been reported in any patient with FGFR associated autosomal dominant craniosynostosis. A more recent hypothesis assumes internal decompression and concomitant external compression of the calvarium as the cause of craniolacunia.2 Accordingly, an abnormal contact between brain and membranous calvarium may cause abnormal radial progression of ossification during early embryogenesis (8-12 weeks’ gestation). Based on the frequent finding of an association of craniolacunia and spina bifida, some authors have suggested that the continued escape of fluid from a myelocele or into an encephalocele might cause microcephaly and craniolacunia.3 However, none of these mechanistic explanations is entirely convincing. As a first step towards an understanding of craniolacunia at the molecular level, we report on a mutation in the fibroblast growth factor receptor 2 (FGFR2) gene in a girl with severe craniolacunia.

The patient was born at term to a 38 year old mother and a 46 year old father after an uneventful pregnancy. At birth, she presented with cloverleaf skull, microcephaly owing to pansynostosis of the calvarial sutures, and proptosis. Magnetic resonance imaging (MRI) showed an Arnold-Chiari malformation. The thumbs and big toes were slightly broadened. No other limb abnormalties were observed. 3D computed tomography of the skull at 3 months showed characteristic craniolacunia (fig 1A-C). She underwent craniectomy at 3 months and resection of the cerebellar tonsils six weeks later. MRI investigation of the vertebral column at 12 months showed normal vertebrae without evidence of spina bifida. At 1 year of age she is developing normally and is starting to walk and talk.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

3D computed tomography of the patient’s skull at 3 months (A-C). (D) DNA sequence of exon 7 showing a G→T transversion at nucleotide position 833 (counted from the first base of the start codon ATG). This results in the exchange of a cysteine for a phenylalanine at amino acid position 278.

EDTA blood from the patient and her clinically normal parents was obtained and DNA was extracted according to standard procedures. Exon 76 (exon U according to Miki et al,5 exon 5 according to Johnsonet al 4) of the FGFR2 gene was amplified by PCR using primers 5′GTCTCTCATTCTCCCATCCC3′ (forward) and 5′GAAGGAGACCCCAGTTGTG3′ (reverse) (30 cycles at 94°C, 61°C, 72°C for 30 seconds each preceded by three minutes at 94°C and followed by a final extension at 72°C for seven minutes). In addition, paternity was confirmed using the STRP markers HUMFIBRA, HUMVWFA31/A, D18S51, HUMTHO1, D19S253, and D21S11 of Urquhart et al.7

The amplification product of exon 7 was sequenced and a G→T transversion was observed at nucleotide position 833 (counted from the first base of the start codon ATG) resulting in the exchange of a cysteine for a phenylalanine at amino acid position 278 in the patient. This cys278phe mutation was not found in either parent. Paternity was confirmed using six STRPs, all of which were informative (not shown), thus indicating that the mutation had occurred de novo. Changes of the cysteine at position 278 have been detected in other cases of craniosynostosis,8-10 but craniolacunia was not reported in any of these. Alteration of the cysteine disrupts a central disulphide bond within immunoglobulin-like loop III of the FGF receptor. Consequently, its structural integrity is destroyed and the receptor might become constitutively activated,11 as suggested by experiments in a Xenopus oocyte system. The formation of novel intermolecular disulphide bonds might cause this activation,12 which appears to trigger premature closure of the calvarial sutures. Although a role of the defective receptor in the origin of craniolacunia can at present not be proven, the physiological function of FGFs in osteogenesis13 makes such a function likely. Given that identical mutations in FGFR can result in completely different phenotypes,14 15 additional genes (“modifier genes”) are thought to play a major role in the resulting phenotype of FGFR associated craniosynostoses. In the present case this phenotype includes craniolacunia. The absence of spina bifida in our patient precludes a mechanistic model involving abnormal flux of cerebrospinal fluid from a myelocele.2