16ournal ofNeurology, Neurosurgery, and Psychiatry 1994;57:1216-1220

Mesencephalic and third ventricle cysts: diagnosisand management in four cases

V Th Ramaekers, J Reul, V Siller, A Thron

AbstractFour infants with obstructive hydro-cephalus caused by space occupying thirdventricle and mesencephalic cysts arereported. Despite immediate shuntinsertion in all patients, there was eitherlack of clinical improvement or late onsetof clinical deterioration. Neuroimaging(CT, MRI, and ventriculography) diag-nosed the presence of non-communicat-ing midline outpouchings of the CSFpathways causing obstruction of aque-ductal CSF flow and brainstem signs.The cysts were of different origin. In onepatient it was caused by a previous thala-mic haemorrhage, in another patient byneonatal Escherichia coli meningoven-triculitis. In two cases with obstructivehydrocephalus at birth, the aetiology isunclear. Direct puncture and drainage ofthe cysts led to clinical improvement.The cysts were poorly visualised on CTand could be misinterpreted as anenlarged third ventricle, simulatingcongenital aqueduct stenosis. Carefulneuroradiological investigation is neces-sary to establish an accurate diagnosisand neurosurgical management. In suchcases with hydrocephalus and persistingventricular enlargement despite shunt-ing, CT ventriculography is a useful tool.

(C Neurol Neurosurg Psychiatry 1994;57:1216-1220)

Department ofPaediatrics,University ofAachen,Pauwelsstrasse 30,D-52057 Aachen,GermanyV Th RamaekersDepartment ofNeuroradiology,University ofAachen,Pauwelsstrasse 30,D52057, Aachen,GermanyJ ReulA Thron

BethlehemKrankenhaus,Stolberg, GermanyV SillerCorrespondence to:Dr V Th Ramaekers,Departnent of Pediatrics,Medizinische Einrichtungender RWTH Aachen,Pauwelsstrasse 30, D-52057Aachen, Germany.Received 1 September 1993and in revised form25 March 1994.Accepted for publication3 May 1994

Obstructive hydrocephalus at birth associatedwith dilatation of the third and lateral ventri-cles most commonly results from aqueductstenosis. In most cases the stenosis is sporadicand non-genetic, but rare X linked cases

occur, often associated with adducted thumbsand with a poor prognosis.1-3 Prenatal infec-tion (toxoplasmosis, cytomegalovirus), post-natally acquired meningitic ependymitis, orintracranial haemorrhage can also cause aque-duct stenosis with obstructive hydrocephalus,but the history and the presence of signs suchas periventricular calcifications or choriore-tinitis can help to differentiate between thevarious conditions.3 Establishing the correctdiagnosis in an affected newborn is importantfor reasons of genetic advice and manage-ment. The presence of gross enlargement ofthe third and lateral ventricles on CT, how-ever, can sometimes obscure the exact under-lying aetiology for obstruction of CSF flow.Particularly, diagnostic difficulties can arise inthe differentiation between aqueduct stenosis,

the presence of cysts within the third ventri-cle, and small cysts located in the pontomes-encephalic region with compression of theaqueduct canal.46 Even suprasellar subarach-noid outpouchings can cause compression onthe floor of the third ventricle or aqueductthereby disturbing normal CSF passage fromthe third ventricle towards the aqueductalcanal.6The present report highlights the clinical

and neuroradiological findings as well as theneurosurgical management in four cases withcongenital and postnatally acquired obstruc-tive hydrocephalus due to blockage of CSFflow by midline cysts or outpouchings ofsuprasellar subarachnoid spaces.

Case reportsCASE 1After a normal pregnancy, the first daughterof healthy parents had to be delivered bycaesarian section because of heart beat de-celeration and craniopelvic disproportion.The neonatal period was uneventful. At theage of nine weeks she was referred to usbecause of accelerated head growth, poorsuckling, and generalised hypotonia. Herhead circumference was enlarged (42 cm;>P97) and the anterior fontanel was fulland tense with separated cranial sutures. Onclinical examination she had a high pitchedcry, poor suckling and swallowing, general-ised hypotonia with normal reflexes, and fail-ure of visual fixation with disconjugate androving eye movements. Obstructive hydro-cephalus was confirmed by cranial ultrasonog-raphy and CT, which showed pronounceddilatation of the third and lateral ventricles aswell as an infratentorial midline cyst (fig 1A).After insertion of a ventriculoperitoneal shuntinto the right frontal horn, the size of the ven-tricles diminished but clinical signs of hypoto-nia, poor feeding, and intermittentdisconjugate upward eye deviation persisted.After one week follow up CT and MRI (fig1C and D) showed persistance of a CSF iso-dense midline cyst with pontomesencephaliclocation. The expanding cyst caused distor-tion of brainstem structures and compressionof the aqueduct. Ventriculography (fig 1B)showed the absence of free communicationbetween the ventricular system and the pon-tomesencephalic cyst. The life threateningbrainstem compression by this expanding cys-tic lesion urged further neurosurgical inter-vention. A communication between the cystand right lateral ventricle was established by

1216

on March 15, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1216 on 1 O

ctober 1994. Dow

nloaded from

Mesencephalic and third ventricle cysts: diagnosis and management in four cases

Figure 1 (case 1) (A) Obstructive hydrocephalus with pronounced dilatation of thesupratentorial ventricles and an infratentorial cystic lesion. (B) Ventriculography clearlyshows the absence ofa free communication between the ventricles and thepontomesencephalic cyst. (C and D) Tl weighted MRI showing a large mesencephalic cystcompressing the aqueduct. A communication to the ventricular system cannot be excluded.

implantation of a cystoventricular shunt.Postoperatively, the patient's conditionimproved rapidly and one month later cranialultrasound showed reduction in size of thepontomesencephalic cyst. Four months laterthe size of the cyst and the third and lateralventricles had reduced considerably. Thefeeding difficulties and roving eye movementsdisappeared and she started to fix and followvisual lures. At the age of 7 months hermotor development was still delayed withsigns of left sided convergent squint andhemiparesis.

CASE 2

After delivery by caesarian section, the patientneeded immediate resuscitation because ofsevere postnatal apnoea and bradycardiaattributed to severe hydrocephalus. His headcircumference was very enlarged (47 cm).Height (52 cm) and weight (3860 g) were

normal. Lumbar puncture pneumoen-cephalography (in 1974) showed normal airfilling of the cervical spinal canal and suboc-cipital and basal cisterns but no filling of theventricular system. Transfontanel punctureshowed a considerably enlarged lateral andthird ventricular system but no visualisation ofthe aqueduct of Sylvius. The presumptive

diagnosis was consistent with congenitalaqueduct stenosis after exclusion of choriore-tinitis and TORCH or Listeria infection and aventriculoatrial shunt was implanted. Clinicalfollow up showed pronounced delay of mentaland motor milestones. By the age of 3 yearsthe patient could walk independently but hisgait was ataxic with signs of spastic paraplegia.His vision was poor with acuity of 2/10 andfunduscopy showed non-progressive bilateraloptic atrophy. He developed peculiar eyemovements consisting of intermittent upwardeye deviation associated with rhythmic hori-zontal head bobbing.7 From 4 years left sidedhemiconvulsions occurred and these weretreated with primidone. At the age of 6 yearsshunt revision was necessary because of acutehydrocephalus (fig 2A and B). Follow up CTat 15 years because of sudden clinical deterio-ration with acute pressure signs indicated thepresence of a large third ventricle cyst block-ing the foramina of Monro (fig 2C).Neurosurgical intervention alleviated theacute pressure signs and was followed byprompt clinical recovery. The cyst was punc-tured and drained by a separate drain thatwas connected to the existing ventricularshunt by a Y connection (fig 2D). Post-operative CT now clearly illustrated callosalagenesis and evidence for pachygyria parti-cularly in the parieto-occipital cortex. Theboy still has severe mental retardation andrecurrent episodes of abnormal upward eyedeviation with preserved consciousness.Although his gait has improved, pyramidalsigns in the lower limbs and a discrete ataxiastill persist.

CASE 3After a gestational period of 35 weekscomplicated by eclampsia, this patient wasdelivered by caesarian section because of fetaldistress. The Apgar score was 8-9-10 (umbil-ical artery pH 7-26). His birth weight was1900 g, length 42 cm, and head circumfer-ence 30 5 cm. Inadequate respirationprompted treatment with oxygen during thefirst hours after birth, on which he recoveredquickly. One mg vitamin K was given intra-muscularly. After an uneventful neonatalperiod during the first 20 days, sudden clinicaldeterioration with seizures and a fall inhaemoglobin content occurred as a conse-quence of a right sided intrathalamic haemor-rhage with intraventricular extension (fig 3A).After development of posthaemorrhagichydrocephalus (fig 3B) an intraventricularRickham reservoir was inserted for regularventricular taps.

Follow up examination by cranial ultra-sound and CT two weeks later showed com-plete resolution of the intrathalamic bleedingsite and after three months, transformationinto a low density area. The borders of thelesion did not show contrast enhancement.Despite regular taps through the Rickhamreservoir, accelerating head growth and raisedintracranial pressure necessitated insertion ofa ventriculoperitoneal shunt at the age of 4months. At the age of 6 months infantile

1217 on M

arch 15, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.10.1216 on 1 October 1994. D

ownloaded from

Ramaekers, Reul, Siller, Thron

.. __ ~~~~~~~~~.... ......:..!i.

Figure 2 (case 2) (A) P-ronounced dilatation of the ventricular system despite a frontalshunt system. The large midline cyst is not identifiable. (B) After shunt revision with a

shunt system placed in the left frontal horn. -Ventricles are still asymmetric, but reduced in

size. (C and D) Nine years later CT shows a small ventricular system without signs ofshunt dysfunction and a large midline cyst. Placement of a second shunt system into thecyst resulted in the collapse of the cyst, which did not communicate with the ventricles.

spasms had to be treated several times withACTH and phenobarbitone until the age of 1year. Follow up examination showed grossdevelopmental delay with a right sided squintand signs of spastic diplegia. At the age of 2years a febrile gastrointestinal infection withSalmonella typhimurium was accompanied bygeneralised seizures 30 minutes in duration.After six days he was transferred to our hospi-tal because of insufficient postictal recovery,intermittent periods of somnolence, andrecurrent vomiting. An EEG showed diffuseslowing with delta activity particularly over

the right hemisphere. Funduscopic examina-tion excluded papilloedema but showed dis-crete pallor of the left optic disc. On CT thethird ventricle appeared grossly enlarged com-pared with its size on previous CT. Contrastinjection into the ventriculoperitoneal shuntshowed normal contrast filling of both lateral

ventricles but a large pear shaped cystoccupying the third ventricle and blockingCSF outflow at both foramina of Monro(fig 3C and D). A catheter was introducedby neurosurgical intervention, passingthrough the left occipital horn and lying withits distal end inside the third ventricle cyst.Only a small wall separated the enlarged thirdventricle cyst and the distended left posteriorlateral ventricle horn, which was close to thebrain surface (fig 3D). So a drain could beplaced easily through the occipital horn pene-trating the thin separating wall without caus-ing much tissue damage and consequentneurological deficits. Postoperatively, theextracranially fixed shunt enabled thedrainage of contents of the third ventricle cystinto the left ventricle, from which CSF wasshunted through the previously implantedventriculoperitoneal shunt. There was promptclinical recovery.

CASE 4At the age of 2 weeks the patient hadEscherichia coli meningitis and subsequentlydeveloped postinfectious hydrocephalus forwhich implantation of a ventriculoperitonealshunt was necessary at the age of 3 months.Neurodevelopmental progress was delayedand he developed cerebral palsy with mentalretardation, spastic tetraplegia, and epilepsyfor which he was put on phenobarbitone.

At the age of 8 years he showed signs ofraised intracranial pressure due to acutehydrocephalus caused by insufficient ven-triculoperitoneal drainage. Preoperative CTshowed dilated lateral ventricles and for thefirst time a non-communicating cystic lesionat the level of the third ventricle was sus-pected (fig 4A).From the age of 8 years he complained sev-

eral times of headaches and nausea, but therewere no signs of raised intracranial pressure.From the age of 13 years these symptomsrelapsed more frequently, and subsided spon-taneously. A CT at that age showed a largemidline cyst bulging into the third ventricle(fig 4B). At 15 years he suddenly developednausea, vomiting, and headaches and becamesomnolent. A new shunt system wasimplanted but he continued to haveheadaches. Ventriculography and CT con-firmed a local expanding cyst within the thirdventricle, which had no free communicationwith the ventricular system (fig 4C and D).

Several neurosurgical procedures wereneeded to reestablish communicationbetween the cyst and lateral ventricles. Twoshunt systems have been inserted draining theleft and right lateral ventricles. To establish afree communication between the third ventri-cle cyst and the lateral ventricles, a stereotacticguided fenestration of the cyst walls was per-formed. The symptoms and signs of raisedintracranial and local pressure had disap-peared at postoperative follow up.

DiscussionDemonstration by CT of obstructive hydro-cephalus with dilatation of the third and

1218

..........

on March 15, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.10.1216 on 1 O

ctober 1994. Dow

nloaded from

Mesencephalic and third ventricle cysts: diagnosis and management in four cases

Figure 3 (case 3) (A) Right sided thalamic bleeding with intraventricular extension. (B)Resorption of the haemorrhage was followed by pronounced dilatation of the ventricularsystem requiring neurosurgical shunting. (C and D) After insertion of a shunt theenlargement of the third ventricle still persists and ventriculography shows a large non-communicating cyst blocking CSF outflow at the foramina ofMonro.

lateral ventricles in the presence of a normalsized fourth ventricle indicates obstruction ofCSF flow through the aqueduct of Sylvius. Inmost congenital cases the underlying cause isunknown and presents a sporadic non-geneticoccurrence, but in a few patients a geneticallyX linked aqueductal stenosis is present.1-3 Insome cases, prenatal infections due tocytomegalovirus and toxoplasmosis or post-natally acquired meningitic ependymitis canalso stenose, distort, or obstruct the aqueduc-tal canal.3 Pitfalls in the establishment of acorrect diagnosis can occur, however; asuprasellar subarachnoid cyst or craniopha-ryngeoma may enlarge and extend into thefloor of the third ventricle to obstruct CSFflow at the foramina of Monro or the exit ofthe third ventricle. The resulting appearanceon CT of obstructive hydrocephalus may some-times simulate stenosis of the aqueduct.5 8In such cases Leo et al have advocated the useof contrast ventriculography because the con-trast medium will not enter the cyst but willenter a dilated third ventricle in a patient withstenosis of the aqueduct.' Kishore et al have

also pointed out the limitation of diagnosis ofintracranial midline cysts (craniopharynge-oma, epidermoid cyst) by CT, particularlythose midline cystic lesions with thin non-contrast enhancing walls containing CSF iso-dense material.4 A further limitation of CT incorrect diagnosis of these lesions is related tovariations in the anatomy at the suprasellarregion and the skull base.9 Binitie et al docu-mented a child with hydrocephalus caused by asuprasellar subarachnoid outpouching thatdistended and filled slowly by a valve mecha-nism. This subsequently invaginated the floorof the third ventricle, and eventually evolvedinto a cystic lesion occupying space at the siteof the third ventricle and obstructing normalCSF flow.6

Therefore, in selected cases a strategyshould be chosen for additional neuroimagingby MRI, ventriculography, or CT metriza-mide cisternography.46As cases 1 and 3 have shown, the persis-

tence of brainstem signs after shunted hydro-cephalus (abnormal eye movements,somnolence, seizures, hypotonia, poor suck-ling, and persistent vomiting) should raisesuspicion as to the presence of an underlyingthird ventricle or brainstem cyst with aque-duct compression and should lead to furtherneuroradiological investigations.4-'0 Amongsuspected cases follow up CT after shuntimplantation should be performed earlybecause visualisation of an enlargement ofmidline CSF pathways can confirm the suspi-cion of the presence of a cyst. Early diagnosisof mesencephalic cysts with obstructivehydrocephalus is important because surgicalintervention with drainage can prevent furtherexpansion of pontomesencephalic cysts andavoid life threatening and irreversible damageto vital brainstem structures. As cases 1, 3,and 4 have illustrated, MRI or contrast ven-triculography through a previously implantedventriculoperitoneal shunt can visualise andoutline the cyst margins and determinewhether it is communicating with the ventric-ular system or not.The underlying aetiology of the mesen-

cephalic cyst in cases 1 and 2 remains unclear,but an arteriovenous malformation or tumourwas ruled out. This was supported by followup neuroimaging that did not show anychange or contrast enhancement by the cystmargins. In cases 2 and 4 the suprasellar sub-arachnoid pouch invaginating the floor of thethird ventricle and eventually causing obstruc-tive hydrocephalus6 could be the possiblepathogenetic factor. A lumbar cisternographymay be the best way to confirm this, but it wasnot done in our patients. If a pouch and not acyst, however, was the underlying aetiology,the neurosurgical treatment would be similar.Moreover, the development of Escherichia colimeningitic ependymitis during the neonatalperiod in case 4 was believed to be responsiblefor the development of a postinfectious cystwithin the ventricular system. In case 3 theunexpected clinical deterioration due to uni-lateral intrathalamic bleeding with intraven-tricular extension at the age of 20 days could

1219 on M

arch 15, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.10.1216 on 1 October 1994. D

ownloaded from

Ramaekers, Reul, Siller, Thron

i_L.I.Figure 4 (case 4) (A) CT shows a cystic formation close to the third ventricle withoutcommunication and hydrocephalus internus despite shunting. (B) Control scans show alarge midline cyst separatedfrom the third ventricle, producing occlusive hydrocephalus.(C) Filling of the cyst with contrast medium shows absence ofcommunication to theventricles. (D) No opacification of the cyst is visible at ventriculography.

not be explained by an arteriovenous malfor-mation. Unilateral thalamic haemorrhagewith intraventricular extension has beenreported in full term infants who were initiallywell but presented with sudden clinical deteri-

oration during the first or second week oflife." 12 An important finding in case 3 was thedocumentation of transformation of the previ-ous intrathalamic bleeding site into a slowlyexpanding third ventricle cyst. The expansionof this only caused clinical signs after a delayof almost two years. A similar course is illus-trated by case 4 with postinfectious hydro-cephalus, in whom expansion of the thirdventricle cyst caused pronounced symptomsafter a delay of 14 years. Successful insertionof a cystoventricular shunt in cases 1 and 3alleviated signs of brainstem dysfunction andlocal pressure.

In conclusion, the advent of additionalMRI and contrast ventriculography can con-firm the correct diagnosis of third ventricle ormesencephalic cysts simulating aqueductstenosis on conventional CT. Early and cor-rect diagnosis is mandatory to take the appro-priate therapeutic steps. Regardless of theaetiology, the neurosurgical treatment can bedifficult and in some cases several procedures(combining shunts and fenestration) may benecessary.

I Bickers DS, Adams RD. Hereditary stenosis of the aque-duct of Sylvius as a cause of congenital hydrocephalus.Brain 1949;72:246.

2 Holmes LB, Nash A, Zurhein GM, Levin M, Opitz JM. Xlinked aqueductal stenosis. Pediatrics 1973;51:697-704.

3 Brett EM. Paediatnic neurology. 2nd ed. Edinburgh:Churchill Livingstone 1991: 467-8.

4 Kishore PRS, Krishna Rao CVG, Williams JP, Vines FS.The limitation of computerized tomographic diagnosisof intracranial midline cysts. Surg Neurol 1980;14:417-31.

5 Leo JS, Pinto RS, Hulvat GF, Epstein F, Kricheff II.Computed tomography of arachnoid cysts. Radiology1 979;130:675-80.

6 Binitie 0, Williams B. Case CP. A suprasellar subarach-noid pouch; aetiological considerations. _7 NeurolNeurosurg Psychiatry 1984;47:1066-74.

7 Nellhaus G. The bobble-head doll syndrome: a "tic" with aneuropathologic basis. Pediatrics 1967;40:250-3.

8 Lee BCP. Intracranial cysts. Radiology 1979;130:667-74.9 Wiggli W, Benz FU. Normal computed tomography

anatomy of the suprasellar subarachnoid space.Radiology 1978;128:65-70.

10 Fenton T, Seton C, Hall D, Kendall B. Intermittent dis-conjugate eye movements-a sign of raised intracranialpressure with brain stem compression. Neuropediatrics1 989;20:30-2.

11 Roland EH, Flodmark 0, Hill A. Thalamic haemorrhagewith intraventricular haemorrhage in the full-term new-born. Pediatrics 1990;85:737-42.

12 Trounce JQ, Fawer CI, Punt J, Dodd KL, Fielder AR,Levene MI. Primary thalamic haemorrhage in the new-born: a new clinical entity. Lancet 1985Ji:190-2.

1 220 on M

arch 15, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.10.1216 on 1 October 1994. D

ownloaded from