Childs Nerv Syst (2005) 21: 889–901DOI 10.1007/s00381-004-1115-z FOCUS SESSION

Giuseppe CinalliPietro SpennatoChristian Sainte-RoseEric ArnaudFerdinando AlibertiFrancis BrunelleEmilio CianciulliDominique Renier

Received: 1 September 2004Published online: 5 May 2005# Springer-Verlag 2005

Chiari malformation in craniosynostosis

Abstract Introduction: Chiari mal-formation (CM) is a frequent findingin multisutural and syndromic cra-niosynostosis, occurring in 70% ofpatients with Crouzon’s syndrome,75% with oxycephaly, 50% withPfeiffer’s syndrome and 100% withthe Kleeblattschädel deformity. Thepathogenesis of this condition andrationale for treatment are still con-troversial. Discussion: Since its firstdescription in 1972, several factorshave been cited to play a role ininducing CM. In the light of recentpublications, the roles of prematurefusion of cranial vault and cranialbase sutures, of congenital anomaliesof the cerebellum and brain stem, ofraised intracranial pressure, of venoushypertension and of hydrocephalusare reviewed. Evaluation and man-agement of CM are also discussed.Conclusion: Chiari malformationappears to be an acquired and pro-gressive condition that develops in thefirst months of life, because of adisproportion between hindbraingrowth and an abnormally small

posterior fossa, a consequence of thepremature fusion of lambdoid andcranial base sutures. Venous hyper-tension caused by stenosis of thejugular foramen can also be present inthese patients, resulting in intracranialhypertension and/or hydrocephalus.Careful MRI evaluation is recom-mended for the forms of craniosyn-ostosis at a high risk of developinghindbrain herniation. The selection ofposterior cranial vault expansion asthe first surgical procedure is advo-cated. In selected cases, treatment ofthe posterior cranial deformity byoccipital vault remodelling and treat-ment of the Chiari-like deformity bysuboccipital decompression can becarried out using the same surgicalprocedure.

Keywords Chiari malformation .Chronic tonsillar herniation .Lambdoid suture . Venoushypertension . Craniosynostosis .Hydrocephalus

G. Cinalli . P. Spennato . F. AlibertiDepartment of Pediatric Neurosurgery,Santobono-PausiliponChildren’s Hospital,Naples, Italy

G. Cinalli (*)Via Gennaro Serra n. 75,80132 Naples, Italye-mail: giuseppe.cinalli@fastwebnet.itTel.: +39-335-6845214Fax: +39-81-2205660

E. CianciulliDepartment of Pediatric Neuroradiology,Santobono-PausiliponChildren’s Hospital,Naples, Italy

C. Sainte-Rose . E. Arnaud . D. RenierDepartment of Pediatric Neurosurgery,Hôpital Necker-Enfants Malades,Paris, France

F. BrunelleDepartment of Pediatric Radiology,Hôpital Necker-Enfants Malades,Paris, France

Introduction

The association between Chiari malformation (CM) andcraniofacial synostosis was first noted by Saldino et al.[53], who in 1972 described a child with Pfeiffer’s syn-drome who, following bicoronal, biparietal and suboccipitalcraniotomies, developed “a communicating hydrocephalusassociated with an Arnold–Chiari malformation and bilat-eral tonsillar herniation”. Until the advent of MRI eval-

uation in pediatric neurosurgical practice, isolated caseshad been reported in the literature [14, 16, 65] and path-ogenesis of the association between craniosynostosis andCM and/or hydrocephalus was only conjectured. Venes[65] described a newborn with Kleeblattschädel deformitywho, 6 weeks after radical craniectomy, developed epi-sodes of respiratory irregularities and bradycardia. MRI,obtained thereafter, showed the presence of hydrocephalusand tonsillar herniation. Frim et al. [16] described a child

with craniofacial dysmorphism and hydrocephalus in whomMRI, performed at 23 months of age to investigate epi-sodes of apnea, revealed herniation of the cerebellar tonsils.This child was managed by wide suboccipital decom-pression. Previous MRI performed when the patient was7 months old was normal, suggesting the acquired nature ofCM in this case. In 1992, Francis et al. [14] reported aseries of ten patients affected by Crouzon’s syndrome, inwhich five had CM. In 1995, we [8] published a report of alarge clinical series based on the MRI evaluation of 95patients with syndromic craniosynostosis, pointing out thehigh frequency and the clinical importance of hindbrainherniation in syndromic craniosynostosis, delineating thepathogenesis and, in a subsequent paper [9], the rationalefor treatment.

Pathogenesis of Chiari malformation (chronic tonsillarherniation, hindbrain herniation) in craniosynostosis

Although Chiari [4] thought that his type I malformationresulted from fetal hydrocephalus with subsequent con-genital tonsillar herniation, many clinical and experimentalstudies indicate that chronic tonsillar herniation observedin CM may result from overcrowding within a primarysmall and shallow posterior cranial fossa due to an under-developed occipital bone [15, 27, 34, 35, 60]. Downwardherniation of neural tissue through the foramen magnummay be secondary to a disproportion between the posteriorfossa and the growing hindbrain structures [34, 60]. In thestudies by Marin-Padilla and Marin-Padilla [27], a highdose of vitamin A given to pregnant hamsters induced inthe offsprings, among other things, underdevelopment ofthe basi-occiput and a small posterior fossa. This resulted incaudal displacement of the cerebellum, which increasedduring its postnatal growth spurt.

Following the observation of CM associated with alumbo-peritoneal shunting procedure [5, 6, 21], increasingattention has been given to acquired CM. This can be as-sociated with several conditions that determine reductionin intracranial volume (cephalocranial disproportion), es-pecially in the posterior fossa, with no associated primarypathological abnormality of the brain, such as the thicken-ing of the cranium secondary to disorders of the bone (os-teopetrosis, fibrous dysplasia) [42] or supratentorial masslesions [31].

Premature fusion of cranial vault and/or cranial base su-tures in craniosynostosis may also be a cause of cephalocra-nial disproportion, leading to overcrowding of the posteriorfossa and hindbrain herniation [8, 65]. Other mechanisms,such as congenital anomalies of the cerebellum and brainstem [2], brain turgor [41], hydrocephalus and venous hy-pertension [14, 18, 61, 62, 68] have been reported to playa role in the pathogenesis of hindbrain hernia. However,the observation that in most cases of craniosynostosis hind-brain herniation is not present at birth, but develops par-

allelling the modification of the skull shape secondary topremature closure of the lambdoid and cranial base sutures(usually between 3 and 6 months of age) further supportsthe pathogenetic hypothesis of overcrowding of the poste-rior fossa secondary to premature sutural fusion [8]. In thispaper, to simplify the terminology of the condition knownwith several synonyms (Chiari, Chiari I, hindbrain her-niation, chronic tonsillar herniation, CM or deformity), inspite of the growing evidence in favor of the acquired na-ture of the condition and in an attempt at avoiding con-fusion by adding new classifications or definitions of whathas already been described and classified 100 years ago, wewill refer to it using the original definition of “Chiarimalformation”.

Genetic alterations

Recent studies have demonstrated that mutations in three ofthe four known fibroblast growth factor receptor (FGFR)genes are responsible for different types of syndromiccraniosynostoses such as Crouzon’s, Apert, Pfeiffer’s andJackson–Weiss syndromes [17, 36, 40]. The genetic mu-tation seems to be located mainly on the FGFR1 gene forPfeiffer’s syndrome, on FGFR2 for Apert, Crouzon’s,Pfeiffer’s and Jackson–Weiss, on FGFR3 for a subtype ofCrouzon’s syndrome associated with acanthosis nigricans.The same FGFR mutations can cause different phenotypes,even different syndromes; at the same time, different mu-tations, even on different genes, can cause the same pheno-type, and the tendency is toward a definition of “nucleotidic”subtypes of the same clinical syndrome that could accountfor the different phenotypic expression and for the differentdegrees of gravity. Some authors [33] tried to find a cor-relation between the mutation observed and the presence ornot of CM. They suggested that in Crouzon’s syndrome,the patients affected by CM and syringomyelia present avariety of mutation that spreads over exons IIIa and IIIc ofthe FGFR2 gene. The involvement of exons IIIa and IIIc ofFGFR2 in Crouzon’s patient with CM was confirmed by asubsequent study [17], which described a novel, previouslyunreported FGFR2 mutation that has been found to beassociated only with Crouzon’s syndrome with Chiari I andsyringomyelia, not to Crouzon’s syndrome without CM.The cascade of events from the nucleotidic mutation to thefinal phenotype is largely unknown and beyond the pur-pose of this paper.

Patterns of growth of the cranial vault and the cranialbase in craniosynostosis

Embryological development of the skull is different for thebones of the vault, in which a membranous ossificationtakes place (desmocranium), from the bones of the cranialbase, in which the ossification is cartilaginous (chondro-

890

cranium) [32]. In the calvaria the growth of the skull issecondary to the displacement of the frontal, parietal andoccipital bones secondary to the rapid growth of the brainin the first years of life, with concomitant osseous de-position at the sutural margins [12]. The growth of thecalvarial sutures follows the brain growth curve. It is ex-tremely rapid during the first few years of life, but slowsdown by the age of 6–7 years [24].

Otherwise, the growth of the cranial base is adapted notonly to calvarial growth, but is also strictly related to thegrowth of the facial skeleton [24, 56]. The growth of theanterior cranial fossa depends primarily on displacementbetween the frontal, sphenoid and ethmoidal bones, oc-curring at the spheno-frontal and spheno-ethmoidal suture.These sutures grow very rapidly in the first few years oflife, but cease growing around the age of 7 years [24]. Themiddle cranial fossa continues to grow several years afterthe growth of the anterior cranial fossa has ceased. Thedisplacement between bones occurs predominantly atspheno-petrosal and spheno-occipital synchondroses [24].The posterior cranial fossa grows in length in early child-hood in the intra-occipital, petro-occipital and spheno-occipital synchondroses. Growth in the intra-occipital ceasesin early childhood, while growth in the spheno-occipitalcontinues past puberty [24].

In syndromic and complex multisutural craniosyn-ostosis, unlike monosutural craniosynostosis, the facialskeleton and the cartilaginous cranial base are primarilyinvolved [24]. In Crouzon’s and Apert syndrome the patternof involvement of the skull base synchondrosis and thetiming of fusion are different. In the Apert syndrome thespheno-occipital, petro-occipital and occipital synchondro-sis are never fused in the first year of life and begin to fusebetween 12 and 48 months to be finally completely fusedafter 4 years of age [25]. In Crouzon’s syndrome thespheno-occipital, petro-occipital and occipital synchondro-sis are in several cases completely fused in the first year oflife [25]. This results in significant differences in the finalanatomy of the skull base, especially concerning the basi-occiput and the posterior cranial fossa. The Apert basi-occiput is larger than normal while that in Crouzon’ssyndrome is smaller [47]. Moreover, the latter preferen-tially expands along a supero-inferior axis whereas little orno growth is allowed along an antero-posterior axis (Fig. 1);the foramen and the basion–opisthion area presents onlysmall changes, whereas the more significant alterations arein the cranial base posterior to the foramen [47], where thecerebellum is located. These differences in growth of theposterior cranial base are equally distributed along differenttime intervals from neonatal age through childhood andadolescence [47]. The interactions of these factors results inaltered dimensions of the posterior cranial fossa: normal orlarger than normal in the Apert syndrome [47], smaller,shortened in the antero-posterior axis and elongated in thesupero-inferior and lateral axis in Crouzon’s syndrome [47,

56]. Moreover, premature fusion of the petro-occipitalsynchondroses, may lead to stenosis or atresia of the jug-ular foramen [51, 52], which is also a common finding inachondroplasia [52, 59].

Patients with oxycephaly and plagiocephaly may havepremature fusion of the cranial base, but not in all cases; inscaphocephaly involvement of the cranial base is excep-tional (Fig. 2) [24]. Nevertheless, recent studies [56] haveshown that cranial base growth is altered not only in syn-dromic or multisutural craniosynostosis, but also in the morefrequent monosutural or bi-sutural craniosynostosis, evenin the absence of primary involvement of cranial base syn-chondrosis. In detail, and whatever the craniosynostosisconsidered, the anterior cranial fossa tends to be underde-veloped in males and overdeveloped in females, the mid-dle fossa tends to be larger than normal in both sexes, theposterior fossa is smaller than normal in both sexes in thefirst 2 years of life, more markedly in women, with smallerantero-posterior dimensions and larger width [56].

The role of the lambdoid suture

In a previous paper, we [8] retrospectively reviewed stan-dard X-ray films, cephalometric radiographs and xerora-diographs of the skull performed in the first 6 years of lifeof children affected by Crouzon’s and the Apert syndromewith the aim of correlating pattern of closure of the suturesand development of CM. Unfortunately, the sutures andsynchondroses of the cranial base are extremely difficult tovisualize on skull radiographs, so we only compared the

Fig. 1 Severe hypoplasia of the posterior fossa in a Crouzon’s patientwith Chiari malformation (CM). Note the preferential expansion ofthe posterior fossa volume along the supero-inferior axis and thesevere obstacle to growth along the antero-posterior axis (arrows),which is especially evident in the posterior half of the posterior fossa.Reprinted and modified with permission from Cinalli et al. [8]

891

patterns of closure of the sagittal and lambdoid suture inpatients demonstrating CM and in patients with normalposterior fossa. However, it is possible that, according toVenes [65], the premature closure of the lambdoid suturereflects the primary closure of the spheno-occipital synchon-drosis. Therefore, the severity and precocity of synostosis ofthe lambdoid suture could be a reliable radiological indicatorof the synostosis of the posterior cranial base synchondro-sis. From this study, it was evident that in Crouzon’s syn-drome the sagittal and lambdoid sutures close very early(median 6 and 21 months respectively), and CM was pres-ent in 72.7% of 44 patients studied by MRI, and in 7 casesassociated with syringomyelia [8]. Moreover, the lambdoidsuture synostosis was observed to occur significantly ear-lier in the cases of Crouzon’s syndrome associated withCM [8] compared with Crouzon’s patients without CM(Fig. 3). On the contrary, in the Apert syndrome the cranialvault synostosis occurs very early for the coronal suture(median 5 months) and significantly later (51 and 60 monthsrespectively) for the sagittal and lambdoid sutures. In thesepatients CM was observed in only 1 out of 51 patients who

underwent MRI [8]. These findings were in perfect agree-ment with those reported from Kreiborg et al. [25] whostudied the progressive closure of cranial vault sutures andskull base synchondroses in the Apert and Crouzon’ssyndrome using 3D CT; unfortunately they did not makeany correlation between the pattern of suture closure andCM.

The interpretation of these data would be that the conflictbetween neural growth and skull growth is most dramatic atthe level of the posterior fossa if the lambdoid suture closesduring the first 2 years of life (Fig. 4); in fact, the cerebellargrowth is especially accelerated compared with the fore-brain and brain stem during these first 2 years. Consequent-ly, a patient with craniosynostosis characterized by an earlyclosure of the lambdoid suture (and of the synchondrosesof the cranial base) would be at a higher risk of CM [8, 25,49].

Further studies demonstrated that [9] the incidence ofCMwas as high as 70% in Crouzon’s syndrome [8, 33, 57],75% in nonsyndromic oxycephaly [46], 50% in Pfeiffer’ssyndrome and 100% in Kleeblattschädel deformity [9]. CMwas also found in other types of syndromic craniosynos-tosis such as Seckel syndrome [22], Antley–Bixler syn-drome [3, 26], Shprintzen–Goldberg syndrome [20], somecases of nonsyndromic complex craniosynostosis involv-ing the lambdoid suture and in some rare cases of scapho-cephaly (Fig. 5) [7]. However, the majority of these casespresent as a common feature the presence of a multi-suturalcraniosynostosis, syndromic or nonsyndromic, with a pre-dominant involvement of the posterior aspect of the skull.Even if the skull base synchondroses are not primarilyinvolved, it is generally recognized that skull vault suturessynostosis alter significantly the growth pattern of thecranial base, which can present hypoplasia or deformationeven in the absence of synchondrosal synostosis [55].Therefore, the succession of events could be as follows:.Inthe first 2 years of life, the progressive fusion of thelambdoid suture (associated or not with closure of cranialbase synchondroses) produces alteration in the skull base(Fig. 4) [25] and stenosis of the jugular foramina (if thepetro-occipital synchondroses are primarily involved). The

Fig. 2 a Nonsyndromicscaphocephaly, b associatedwith CM

Fig. 3 In Crouzon’s syndrome, patients affected by CM (dottedline) demonstrate lambdoid suture synostosis in the first 2 years oflife, significantly earlier than patients without CM (continuous line).Reprinted and modified from Cinalli et al. [8] with permission

892

first result would be a small posterior fossa [62] with con-sequent herniation of the cerebellum into the cervical canalduring the phase of rapid neural growth in the very firstmonths of life (Fig. 6) [8]. The second result would bevenous hypertension, induced both by jugular foramenstenosis (if the petro-occipital synchondroses are involved;Figs. 7, 8) and crowding of the posterior fossa with con-sequent compression of the sigmoid sinus [23, 28]. Thesefactors can alter the CSF hydrodynamics, impairing CSFcirculation at the level of the posterior fossa with an ob-structive mechanism and impairing CSF reabsorption at the

level of the pacchionian granulations, with the overall finalresult of an increased CSF outflow resistance.

In cases of severe craniosynostosis, such as Kleeblattschä-del deformity, marked changes in the posterior fossa can befound. The severe crowding in the foramen magnum mayresult not only in hindbrain herniation, but also in brain stemcompression and deformation of the fourth ventricle. Thus,hindbrain herniation can be considered to be a conditioncreating or aggravating a hydrocephalic state, not the con-sequence of hydrocephalus. This explains the cases of CMwithout hydrocephalus, more frequently observed in cranio-synostosis without primary involvement of the skull basesynchondrosis (i.e. oxycephaly).

In conclusion, the interactions between the synostosis ofthe lambdoid suture and skull base synchondrosis, thecrowding of the posterior fossa, the jugular foramen ste-nosis with venous sinus hypertension and the increasedCSF outflow resistance give origin to complex pathophys-iological mechanisms inducing CM alone or CM and hy-drocephalus (Fig. 9).

The role of sagittal, coronal and metopic sutures

In case of very early closure of sagittal and coronal sutures(in utero), a cephalocranial disproportion in the supraten-torial compartment occurs early, forcing the neural growthto be directed posteriorly and inferiorly, pushing down thetentorium. This can induce a lower attachment of the ten-torium, as can be observed in more severe cases of cranio-synostosis. The low attachment of the tentorium, near theforamen magnum, reduces the size of the posterior fossa,increasing the risk of CM, especially if premature lambdoidsynostosis also occurs [8]. We have observed sporadiccases of nonsyndromic scaphocephaly associated with CM(Fig. 2).

Fig. 5 Patient affected by Jackson–Weiss syndrome. Note the pro-lapse of the cerebellar tonsils

Fig. 4 a–c Skull X-ray of a Crouzon’s patient at the age of a 1 week, b 2 months and c 5 months. Note the progressive synostosis of thecoronal and lambdoid suture and the progressive increase in fingerprint impressions

893

The precocity of coronal and sagittal suture synostosis doesnot play a role in the pathophysiology of CM in Crouzon’ssyndrome [8]. In the Apert syndrome the sagittal sutureis often wide open and larger than normal [25]. This spec-ificity of the Apert sagittal suture, allowing a good adap-tation of the interparietal diameter to the brachycephalyinduced by the coronal synostosis, may play a role in thearchitecture of the Apert posterior fossa, which is usuallylarger than normal [47, 48] and is rarely associated with CM[8, 45].

Tubbs et al. [64] reported an incidence of Chiari type Imalformation associated with simple metopic ridges withno other clinical signs of trigonocephaly (i.e. hypotelorism)

of 30%. They hypothesized that a small decrease in anteriorfossa volume in children with simple metopic ridges and noother signs of trigonocephaly may be directly or indirectlyresponsible for a higher incidence of CM in these children.

Congenital anomalies of the cerebellumand brain stem

Since the earliest reports of CM in craniosynostosis, theorigin of this anomaly (congenital or acquired) has beendebated [65]. Radiological and anatomical findings, suchas a straightened brain stem with loss of cephalic and pon-tine flexures, elongation of the brain stem in both upward

Fig. 7 Sinusography in a 3-month-old Crouzon’s patient affected bysignificant multisutural craniosynostosis. Note the stenosis of thejugular foramens and the significant collateral circulation goingfrom the sagittal sinus to the subcutaneous veins

Fig. 8 Angio-MRI of the posterior fossa in a 10-year-old Crouzon’spatient with CM and hydrocephalus. Note the stenosis of theterminal part of the sigmoid sinuses (long arrows) and the collateralcirculation at the level of the mastoid emissary veins (arrowheads)

Fig. 6 Patient affected by theKleeblattschädel deformity.Note the absence of CM at theage of 1 day with a a smallcerebellum, and b cerebellarexpansion at the age of2 months, partially directedinto the upper cervical canal.Reprinted from Cinalli et al.[9] with permission

894

and downward directions, a beaked tectal plate and avertical tentorium with a low insertion, strongly suggesteda primary neural developmental anomaly [2]. These find-ings are usually associated with Chiari II malformationrather than Chiari I, which consists of cerebellar tonsillarherniation alone [16, 65]. However, the other anomaliesassociated with Chiari II malformation and myelodyspla-sia, such as medullary kinking, polymicrogyria, enlargedmassa intermedia and cerebellar dysmorphism, are usuallyabsent in craniosynostosis, unless associated myelomenin-gocele and neural tube defects are present [16, 66]. Theobservations of Hoffman and Tucker [21] and Chumas etal. [5, 6] of acquired CM following lumbo-peritoneal shunt-ing, in which the hindbrain herniation and mesencephalicchanges represented not only simple tonsillar herniation,but also an alteration in the shape of the tectal plate, dis-tortion of the midbrain and obliteration of the mesence-phalic cisterns, may support the acquired origin of theanatomical changes found in CM associated with cranio-synostosis. If crowding forces associated with cephalocra-nial disproportion seen in lumbo-peritoneal shunting cancause anatomical changes resembling both Chiari type Iand Chiari type II malformations, it is reasonable that com-plex crowding forces, such as those of complex multisuturalsynostosis, might produce more complicated anatomicalchanges [16]. Frim et al. [16] suggested that the changesobserved in Chiari type II might be a final common patternfor any posterior fossa compression syndrome, if severeenough.

In conclusion, the role of congenital anomalies of thecerebellum and brain stem in the pathogenesis of CM incraniosynostosis appears not to be crucial. Moreover, in theApert syndrome, the syndromic form of craniosynostosis

more often associated with neural developmental anoma-lies, CM is very rare [8, 43].

The role of hydrocephalus

The association between hydrocephalus and craniosynos-tosis has been well documented and the incidence of hy-drocephalus has been found to be between 4 and 10% [10,11, 13, 14, 19, 37, 52, 58, 62]. The incidence is signif-icantly higher for syndromic craniosynostosis than fornonsyndromic craniosynostosis [10]. In syndromic cranio-synostosis, CM is observed in the majority (88%) of childrenaffected by hydrocephalus [10]. Specifically in Crouzon’ssyndrome, all patients affected by hydrocephalus had CMonMRI, whereas 53% of the patients with CM did not havehydrocephalus [8]. These findings were confirmed by othersubsequent studies [51]. The high incidence of CM in chil-dren with hydrocephalus raises the question about the re-lationship between these two entities in craniosynostosis.These data seem to show that CM is a condition nec-essary but not sufficient for the onset of hydrocephalus.For the same reasons, CM cannot be considered to be thesimple result of chronic intracranial hypertension inducedby hydrocephalus.

The role of venous hypertension

Venous sinus hypertension has been evoked as a possiblepathophysiological factor in the origin of CM. The drivingforce for CSF reabsorption is the difference between CSFpressure and sagittal sinus pressure. Venous hypertension

Fig. 9 Cascade of events lead-ing to CM and hydrocephalus incraniosynostosis

895

induced by jugular foramen stenosis, increasing the sagittalsinus pressure, results in a higher CSF pressure beingrequested to maintain CSF balance [14, 52]. The effect ofthis depends on two main factors: the degree of jugularforamen stenosis [50] and effectiveness of collateral circula-tion [51] and the degree of cranial and brain compliance. Ithas been proposed that the degree of jugular foramen ste-nosis depends on the molecular mutation responsible forthe syndrome: the stenosis (or atresia) would be moresevere in patients carrying the FGFR3 ala391glu muta-tion and less severe in patients with the FGFR2 mutation[29, 33, 50].

In children with closed sutures, intracranial pressure mayrise to very high levels, overcoming the high sagittal sinuspressure; this permits absorption of CSF, with normal-sizedor small ventricles, as seen in some cases of pseudotumorcerebri (Fig. 10a) [39, 52].

In contrast, in infants and children with open sutures (orin craniosynostotic patients following cranial suture release),increased CSF pressure induces progressive head enlarge-ment and dilatation of the ventricles and subarachnoidspaces (Fig. 10b) [10, 52].

Usually collateral venous pathways, through the fora-men magnum and/or through emissary veins and scalpveins, progressively develop and allow a new status quo todevelop (Fig. 7) [10, 50–52, 61]. If the venous obstructionis not compensated the rise in CSF pressure may result inprogressive hydrocephalus and intracranial hypertension.

Francis et al. [14], in their report on chronic tonsillarherniation (CTH) and Crouzon’s syndrome, tried to give aunifying explanation for the co-existence of jugular fora-

men stenosis and CM in craniosynostosis, and focused onjugular foramen stenosis and the resultant increased sagittalsinus pressure and venous turgor of the brain as the primarypathogenetic event. Usually, hydrocephalus occurs in thesepatients following release of synostotic sutures. After shuntinsertion, hydrocephalus resolves and the skull fuses, butincreased venous turgor remains. The increased venousturgor produces a pseudotumor-like syndrome that maylead to the development of CM. This theory was supportedby the mathematical model of ventricular volume regula-tion developed by Rekate [41]. He introduced a mathemat-ical term called kb that described an intrinsic property of theliving brain related to the energy required to distort thebrain itself. A high value of kb indicated incompressibilityof the brain (brain turgor). The increased venous turgor mayelevate both kb and intracranial hypertension, providing thedriving force for tonsillar herniation.

Against this interesting theory still remain the over-whelming data coming from clinical observation, provingthat hydrocephalus is not a necessary factor in the patho-genesis of CM and that in hydrocephalic patients CM isalready present at the time of the diagnosis of hydroceph-alus, prior to shunt insertion. Moreover, CM is very fre-quent in patients with craniosynostosis who do not presentwith primary involvement of skull base synchondrosis (i.e.oxycephaly), therefore without jugular foramen stenosis[46]. Finally, there are sporadic observations of hydro-cephalus in the presence of CM cured by endoscopic thirdventriculostomy followed by resolution of CM that are infavor of an obstructive role of the CM in the pathogenesisof hydrocephalus.

Fig. 10 Multisutural craniosyn-ostosis with CM and bilateralstenosis of the jugular foramen.a Before cranial vault remodel-ling, note the small ventricles,and b after cranial vault remod-elling, note the significant,progressive ventriculardilatation

896

However, other factors have been described to be as-sociated with the development of hydrocephalus: basilarinvagination [1], aqueductal stenosis (Fig. 11) [14] andcompression by the midline occipital bone crest [63].

Evaluation and management

More than one third of patients with hindbrain herniationbecome symptomatic for CTH or develop syringomyeliccavities [8]. Symptoms range from suboccipital pain to life-

threatening brain stem dysfunction. Usually, the onset ofsymptoms occurs later in life. Milhorat [30] reported that20% of patients with Crouzon’s syndrome had developedsymptomatic CTH before the age of 20. However, espe-cially in very young children, the onset of symptoms maybe dramatic, with appearance of respiratory problems suchas central apnea, bilateral vocal cord paralysis, bulbar palsy,ventilatory control abnormalities, persistent cyanosis andbreath-holding spells. These findings indicate the need fora careful neuroradiological follow-up in patients with syn-dromic or complex craniosynostosis to assess the presenceand the evolution of CTH, and raise the question of whetheror not a “prophylactic” posterior fossa decompression shouldbe indicated in asymptomatic babies, at the time of cranialvault remodelling.

Magnetic resonance imaging is the gold standard forevaluation of CTH [8]: the descent of the tonsils can bemeasured by callipers on the sagittal MR images from aline joining the basion to the opisthion; the presence of thecisterna magna and the morphology of the brain stem canalso be assessed. Findings of absence of cisterna magna,asymmetry or displacement of brain stem and cerebellartonsils >2 mm below the basion–opisthion line, are diag-nostic for CTH (Fig. 12).

The patients, who probably exhibit CTH on MRI, areusually candidates for total calvarial reconstruction to cor-rect the significant deformity secondary to bicoronal andbilambdoid synostosis. When both fronto-orbital advance-ment and occipital remodelling are requested, one-step ormultiple-step procedures can be employed. The one-stepprocedure is performed with the child in the “modifiedprone position”, described by Pollack et al. [38]. This al-lows exposure from the orbital ridge to the foramen mag-num, but requires marked hyperextension of the neck. This

Fig. 11 Crouzon’s patient with large posterior fossa and normalcerebellar anatomy affected by hydrocephalus and aqueductal stenosis

Fig. 12 a, b Two examples ofasymptomatic CM in Crouzon’spatients

897

position is contraindicated for patients who present withanomalies of the craniovertebral junction, because the hyper-extending the neck for several hours [38] may lead to pro-longed severe compression on the spinal cord and medulla.Thus, MRI scans should be obtained preoperatively for allpatients affected by syndromic or complex craniosynosto-sis who are candidates for total calvarial reconstruction inthe modified prone position. This MRI examination shouldbe recent enough to exclude possible modifications of theanatomic conditions at the level of the cervico-medullaryjunction due to the progressive nature of CTH, especially inthe first months of life. If CTH is present, the modifiedprone position is contraindicated and a “staged” repair ofthe malformation is advocated [44, 54, 67]. In agreementwith the Craniofacial School of Birmingham [54, 67], thepolicy of early remodelling of the posterior cranium, fol-lowed by anterior (orbitofrontal) advancement, should beadopted in these patients. Early correction of the occipitalflattening allows the release of intracranial pressure anddefers the fronto-orbital advancement, reducing the occur-rence of recurrent craniosynostosis, which would requirelate re-operation [44].

In case of newborns with CM, severe flattening of theposterior fossa and overcrowding of the foramen magnum,a surgical technique to correct both occipital vault flatten-ing and CM during the same procedure has been proposed[9]. The patient is positioned prone with the neck in mildflexion; a biparietal skin incision, starting behind the ear, istraced. The scalp flap is reflected posteriorly to the pro-tuberantia occipitalis externa, where occipital musculatureis elevated subperiostally to allow complete exposure ofthe posterior ridge of the foramen magnum. After removalof the parietal bones and the supratentorial occipital bone,the torcular and transverse sinus are dissected free fromoverlying bone, and epidural dissection is performed in theposterior fossa to the posterior ridge of the foramenmagnum, which must be carefully dissected from the perios-teum. “En bloc” craniectomy of the infratentorial occipitalbone is then possible and the occipital foramen is openedwide laterally. The decompression must be performed aroundthe lateral aspects of the foramen to avoid long-term failuresecondary to the re-growth of bone. The infratentorial com-partment is left uncovered and a bone bridge is built withthe largest and most solid fragments of the previouslydissected cranial vault and securely fixed above the trans-verse sinus. To reshape and lower the vault and induceposterior expansion of the brain, a vycril net is placed onthe bregma and securely fixed bilaterally under tension tothe temporal muscle aponeurosis. The remaining bone frag-ments are then positioned without fixation on the occipito-parietal convexity. In the immediate postoperative periodpatients lie in the lateral position for a few days. Usually,dural opening and laminectomy of the involved cervicallevels are not performed during the procedure. The morpho-logical improvement obtained with this procedure is usuallygood: even if some degree of tonsillar herniation still re-

mains, crowding of the foramen magnum is reduced andmore CSF can be observed around the medulla on post-operative MRI. Follow-up of patients treated according tothis policy is promising [9].

Dura opening and cervical laminectomy, which are in-dicated in cases of severe compression of the medulla (suchas in some cases of Kleeblattschädel), carry the risk of severebleeding from the dural edges, because of the significantcollateral circulation at the level of the foramen magnum,which is present in cases of jugular foramen stenosis. Ac-tually, when venous outflow is severely impaired, sub-cutaneous dissection, due to the presence of a collateralsystem of drainage through the scalp veins, and bone flapremoval, due to the presence of transosseous venous chan-nel in the occipital bone, may also cause troublesome bleed-ing during surgery. A fatal case of acute onset of intracranialhypertension, following surgical interruption of this sys-tem, has been described [63].

When CM is diagnosed or becomes symptomatic later inlife, during childhood or early adulthood, it should bemanaged like any other CM requiring surgical treatment,with the same indications for surgical treatment or simpleradiological follow-up as in any noncraniosynostosis pa-tient. The only major difference is in the preoperative radio-logical evaluation: patients with CM and craniosynostosis,especially if syndromic or associated with hydrocephalus,are likely to present with abnormal venous drainage, whichmust be correctly studied at least with a preoperative angio-MRI. If this examination is positive, the indication for atraditional angiographic study should be seriously takeninto consideration to be perfectly aware of the anatomy ofthe venous system before the surgical approach to the pos-terior fossa.

Conclusions

Chiari malformation is a frequent finding in syndromic andmultisuture craniosynostosis, characterized by early (in thefirst 2 years) fusion of lambdoid sutures and cranial basesynchondroses, such as Crouzon’s syndrome and theKleeblattschädel deformity. CM can be an acquired andprogressive condition that develops in the first months oflife, because of disproportion between hindbrain growthand an abnormally small posterior fossa. Venous hyperten-sion, caused by stenosis of the jugular foramen (secondaryto premature synchondroses of the cranial base bones) canalso be present in these patients, resulting in intracranialhypertension and/or hydrocephalus, even if overcrowding ofthe posterior fossa and CM itself may play an important rolein inducing hydrocephalus.

Patients with CM can remain asymptomatic throughoutlife; however, more than one third of patients becomesymptomatic or develop a syringomyelic cavity. Symp-toms, even if occurring later in life, can be life-threatening,especially if they occur in very young children.

898

The existence of CM contraindicates surgical correctionof the cranial deformity with a single procedure in themodified prone position because of prolonged neck hyper-extension during surgery. We support the selection of pos-terior cranial vault expansion as first surgical procedure torelease intracranial hypertension and delay fronto-orbitaladvancement. For selected cases occurring in the first yearof life, with severe flattening of the posterior fossa, over-crowding of the foramen magnum and CM, treatment ofthe posterior cranial deformity by occipital vault remodel-

ling and treatment of the Chiari-like deformity by suboc-cipital decompression with the same procedure can betaken into consideration. In older patients, especially insyndromic craniosynostotic and/or hydrocephalic patients,preoperative planning must include an angiographic studyof the venous circulation of the posterior fossa.

Acknowledgements All the patients described in this paper wereevaluated and treated in the Department of Pediatric Neurosurgery ofthe Hôpital Necker-Enfants Malades, Paris, France.

References

1. Angle CR, McIntire MS, Moor R(1976) Cloverleaf skull Kleeblattschädeldeformity syndrome. Am J Dis Child114:198–202

2. Carmel PW (1988) Comment on VenesJL: Arnold–Chiari malformation in aninfant with Kleeblattschädel: anacquired deformity? Neurosurgery23:362

3. Chang YT, Tsai FJ, Shen WC, Lin HC,Peng CT, Tsai CH (2000) Antley–Bixler syndrome associated withArnold–Chiari malformation. ActaPaediatr 89:737–739

4. Chiari H (1891) Über Veränderungendes Kleinhirns infolge von Hydroce-phalie des Grosshirns. Dtsch MedWochenschr 17:1172–1175

5. Chumas PD, Drake JM, Del Bigio MR(1992) Death from chronic tonsillarherniation in a patient with lumboperi-toneal shunt and Crouzon’s disease.Br J Neurosurg 6:595–599

6. Chumas PD, Armstrong DC, DrakeJM, Kulkarni AV, Hoffman HJ,Humphreys RP, Rutka JT, Hendrick EB(1993) Tonsillar herniation: the rulerather than the exception after lumb-operitoneal shunting in the pediatricpopulation. J Neurosurg 78:568–573

7. Chumas PD, Cinalli G, Arnaud E,Marchac D, Renier D (1997) Classifi-cation of previously unclassified casesof craniosynostosis. J Neurosurg86:177–181

8. Cinalli G, Renier D, Sebag G,Sainte-Rose C, Arnaud E, Pierre-KahnA (1995) Chronic tonsillar herniation inCrouzon and Apert syndrome: the roleof the premature synostosis of thelambdoid suture. J Neurosurg 83:575–582

9. Cinalli G, Chumas P, Arnaud E,Sainte-Rose C, Renier D (1998)Occipital remodeling and suboccipitaldecompression in severe craniosynos-tosis associated with tonsillar hernia-tion. Neurosurgery 42:66–73

10. Cinalli G, Sainte-Rose C, Kollar E,Arnaud E, Renier D (1998) Hydro-cephalus and craniosynostosis.J Neurosurg 88:209–214

11. Collmann H, Sörensen N, Krauss J,Muhling J (1988) Hydrocephalus incraniosynostosis. Childs Nerv Syst4:279–285

12. Enlow DH (1975) Handbook of facialgrowth. Saunders, Philadelphia

13. Fishman MA, Hogan GR, Dodge PR(1971) The concurrence of hydroceph-alus and craniosynostosis. J Neurosurg34:621–629

14. Francis PM, Beals S, Rekate HL,Pittmann HW, Manwaring K, Reiff J(1992) Chronic tonsillar herniation andCrouzon’s syndrome. Pediatr Neuro-surg 18:202–206

15. Friede RL, Roesmann U (1989)Chronic tonsillar herniation. Anattempt at classifying chronic hernia-tion at the foramen magnum. ActaNeuropathol 34:214–235

16. Frim DM, Jones D, Goumnerova L(1990) Development of symptomaticChiari malformation in a child withcraniofacial dysmorphism. PediatrNeurosurg 16:228–231

17. Fujisawa H, Hasegawa M, Kida S,Yamashita J (2002) A novel fibroblastgrowth factor receptor 2 mutation inCrouzon syndrome associated withChiari type I malformation and syrin-gomyelia. J Neurosurg 97:396–400

18. Girard N, Lasjaunias P, Taylor W(1994) Reversible tonsillar prolapse invein of Galen aneurysmal malforma-tions: report of eight cases and patho-physiological hypothesis. Childs NervSyst 10:141–147

19. Golabi M, Edwards MSB, OusterhoutDK (1987) Craniosynostosis and hy-drocephalus. Neurosurgery 21:63–67

20. Greally MT, Carey JC, Milewicz DM,Hudgins L, Goldberg RB, ShprintzenRJ, Cousineau AJ, Smith WL Jr,Judisch GF, Hanson JW (1998)Shprintzen-Goldberg syndrome: a clin-ical analysis. Am J Med Genet 19:202–212

21. Hoffmann HJ, Tucker WS (1976)Cephalocranial disproportion. A com-plication of the treatment of hydro-cephalus in children. Childs Brain2:167–176

22. Hopkins TE, Haines SJ (2003) Rapiddevelopment of Chiari I malformationin an infant with Seckel syndrome andcraniosynostosis. Case report andreview of the literature. J Neurosurg98:1113–1115

23. Kinal ME (1962) Hydrocephalus andthe dural venous sinuses. J Neurosurg19:195–201

899

24. Kreiborg S (1986) Postnatal growth anddevelopment of the craniofacial com-plex in premature craniosynostosis. In:Cohen MM Jr (ed) Craniosynostosis:diagnosis, evaluation, and manage-ment. Raven, New York, pp 157–189

25. Kreiborg S, Marsh JL, Cohen MM Jr,Liversage M, Pedersen H, Skovby F,Borgesen SE, Vannier MW (1993)Comparative three-dimensional analy-sis of CT-scans of the calvaria andcranial base in Apert and Crouzonsyndromes.J Craniomaxillofac Surg 21:181–188

26. Lee HJ, Cho DY, Tsai FJ, Shen WC(2001) Antley-Bixler syndrome,description of two new cases andreview of the literature. PediatrNeurosurg 34:33–39

27. Marin-Padilla M, Marin-Padilla TM(1981) Morphogenesis of experimen-tally induced Arnold-Chiari malforma-tion. J Neurol Sci 50:29–55

28. Martins AN, Kobrine AI, Larsen DF(1974) Pressure in the sagittal sinusduring intracranial hypertension inman. J Neurosurg 40:603–608

29. Meyers GA, Orlow SJ, Munro JR et al(1995) Fibroblast growth factor recep-tor 3 (FGFR3) transmembrane mutationin Crouzon syndrome with acanthosisnigricans. Nat Genet 11:462–464

30. Milhorat TH (1998) Comment in:Cinalli G, Chumas P, Arnaud E,Sainte-Rose C, Renier D: Occipitalremodeling and suboccipital decom-pression in severe craniosynostosisassociated with tonsillar herniation.Neurosurgery 42:72

31. Morioka T, Shono T, Nishio S, YoshidaK, Hasuo K, Fukui M (1995) AcquiredChiari I malformation and syringomy-elia associated with bilateral chronicsubdural hematoma. J Neurosurg83:556–558

32. Müller F, O’Rahilly (1980) The humanchondrocranium at the end of theembryonic period, proper, with partic-ular reference to the nervous system.Am J Anat 159:33–58

33. Mulliken JB, Steinberger D, Kunze Set al (1999) Molecular diagnosis ofbilateral bicoronal synostosis. PlastReconstr Surg 104:1603–1615

34. Nishikawa M, Sakamoto H, Hakuba A,Nakanishi N, Inoue Y (1997) Patho-genesis of Chiari malformation:a morphometric study of the posteriorcranial fossa. J Neurosurg 86:40–47

35. Nyland H, Krogness KG (1978) Size ofthe posterior fossa in Chiari type 1malformations in adults. Acta Neuro-chir 40:233–242

36. Park WJ, Meyers GA, Li X (1995)Novel FGFR2 mutations in Crouzonand Jackson–Weiss syndromes showallelic heterogeneity and phenotypicvariability. Hum Mol Genet 4:1229–1233

37. Pollack IF, Losken HW, Biglan AW(1996) Incidence of increased intracra-nial pressure after early surgical treat-ment of syndromic craniosynostosis.Pediatr Neurosurg 24:202–209

38. Pollack IF, Losken HW, Hurwitz DJ(1996) A combined frontoorbital andoccipital advancement technique foruse in total calvarial reconstruction.J Neurosurg 84:424–429

39. Ray BS, Dunbar HS (1951) Thrombo-sis of the dural venous sinuses as acause of “pseudotumor cerebri”. AnnSurg 134:376–386

40. Reardon W, Winter RM, Rutland P(1994) Mutations in the fibroblastgrowth factor receptor 2 gene causeCrouzon syndrome. Nat Genet 8:98–103

41. Rekate HL (1992) Brain turgor (Kb):intrinsic property of the brain to resistdistortion. Pediatr Neurosurg 18:257–262

42. Rengachary SS, Blount J, Heros D,Bowers S, Truwit C (1997) Cranioce-phalic disproportion with increased in-tracranial pressure and brain herniation:a new clinical syndrome in anemicpatients: report of two cases. Neuro-surgery 41:297–304

43. Renier D, Sainte-Rose C, Marchac D,Hirsch JF (1982) Intracranial pressurein craniosynostosis. J Neurosurg57:370–377

44. Renier D, Sainte-Rose G, Sainte-RoseC, Marchac D (1995) Recurrent cra-niosynostosis: when and why? Pre-sented at the Consensus Conference onPediatric Neurosurgery: Craniosynos-tosis 95, Rome, 3–6 May 1995

45. Renier D, Arnaud E, Cinalli G, SebagG, Zerah M, Marchac D (1996) Prog-nosis for mental function in Apert’ssyndrome. J Neurosurg 85:66–72

46. Renier D, Cinalli G, Lajeunie E,Arnoud E, Marchac D (1997)L’oxycéphalie, une craniosténosesévère. A propos d’une série de129 cas. Arch Pediatr 4:722–729

47. Richtsmeier JT (1987) Comparativestudy of normal, Crouzon and Apertcraniofacial morphology using finiteelement scaling analysis. Am J PhysAnthropol 74:473–493

48. Richtsmeier JT (1988) Craniofacialgrowth in Apert syndrome as measuredby finite-element scaling analysis. ActaAnat 133:50–56

49. Richtsmeier JT, Lele S (1990) Analysisof craniofacial growth in Crouzonsyndrome using landmark data. JCraniofac Genet Dev Biol 10:39–62

50. Robson CD, Mulliken JB, RobertsonRL, Proctor MR, Steinberger D, BarnesPD, McFarren A, Muller U,Zurakowski D (2000) Prominent basalemissary foramina in syndromiccraniosynostosis: correlation withphenotypic and molecular diagnoses.Am J Neuroradiol 21:1707–1717

51. Rollins N, Booth T, Shapiro K (2000)MR venography in children with com-plex craniosynostosis. Pediatr Neuro-surg 32:308–315

52. Sainte-Rose C, LaCombe J,Pierre-Khan A, Renier D, Hirsh JF(1984) Intracranial venous sinushypertension: cause or consequenceof hydrocephalus in infants? J Neuro-surg 60:727–736

53. Saldino RM, Steinbach HL, Epstein CJ(1972) Familial acrocephalosyndactyly(Pfeiffer syndrome). Am J Roentgenol116:609–622

54. Sgouros S, Goldin JH, Hockley AD,Wake MJC (1996) Posterior skull sur-gery in craniosynostosis. Childs NervSyst 12:727–733

55. Sgouros S, Natarajan K, Hockley AD,Goldin JH, Wake M (1999) Skull basegrowth in childhood. Pediatr Neurosurg31:259–268

56. Sgouros S, Natarajan K, Hockley AD,Goldin JH, Wake M (1999) Skull basegrowth in craniosynostosis. PediatrNeurosurg 31:281–293

57. Shigeta H, Sakai K (1996) [Chiarimalformation (chronic tonsillar hernia-tion) and syringomyelia in Crouzon’ssyndrome] (in Japanese). Nerv SystChild 21:395–401

58. Shiroyama Y, Ito H, Yamashita T et al(1991) The relationship of cloverleafskull syndrome to hydrocephalus.Childs Nerv Syst 7:382–385

59. Steinbok P, Hall J, Flodmark OP (1989)Hydrocephalus in achondroplasia: thepossible role of intracranial venoushypertension. J Neurosurg 71:42–48

60. Stovner LJ, Bergan U, Nilsen G,Sjaastad O (1993) Posterior cranialfossa dimensions in the Chiari I mal-formation: relation to the pathogenesisand clinical presentation. Neuroradiol-ogy 35:113–118

61. Taylor WJ, Hayward RD, Lasjaunias P,Britto JA, Thompson DNP, Jones BM,Evans RD (2001) Enigma of raisedintracranial pressure in patients withcomplex craniosynostosis: the role ofabnormal intracranial venous drainage.J Neurosurg 94:377–385

900

62. Thompson DN, Harkness W, JonesBM, Hayward RD (1997) Aetiology ofherniation of the hindbrain in cranio-synostosis. An investigation incorpo-rating intracranial pressure monitoringand magnetic resonance imaging.Pediatr Neurosurg 26:288–295

63. Thompson DNP, Hayward RD,Harkness WJ, Bingham RM, Jones BM(1995) Lessons from a case of klee-blattschädel. Case report. J Neurosurg82:1071–1074

64. Tubbs RS, Elton S, Blount JP, OakesWJ (2001) Preliminary observations onthe association between simple metopicridging in children without trigono-cephaly and the Chiari I malformation.Pediatr Neurosurg 35:136–139

65. Venes JL (1988) Arnold-Chiari mal-formation in an infant with Klee-blattschädel: an acquired deformity?Neurosurgery 23:360–362

66. Venes JL (1998) Comment in: CinalliG, Chumas P, Arnaud E, Sainte-RoseC, Renier D: Occipital remodeling andsuboccipital decompression in severecraniosynostosis associated withtonsillar herniation. Neurosurgery42:71–72

67. Wall SA, Goldin JH, Hockley AD,Wake MJC, Poole MD, Briggs M(1994) Fronto-orbital re-operation incraniosynostosis. Br J Plast Surg47:180–184

68. Zerah M, Garcia-Monaco M, RodeschG, Terbrugge K, Tardieu M, de VictorD, Lasjaunias P (1992) Hydrodynamicsin vein of Galen malformations. ChildsNerv Syst 8:111–117

901