92ournal ofNeurology, Neurosurgery, and Psychiatry 1993;56:932-935

LESSON OF THE MONTH

Syndrome of transtentorial herniation: is verticaldisplacement necessary?

Allan H Ropper

AbstractMRI from a comatose patient with amassive acute subdural haematomashowed most of the features of transten-torial herniation described in the classicpathology literature. In addition toencroachment on the perimesencephaliccisterns, infarction in the anterior andposterior cerebral artery territories,ischaemic change in the lower dien-cephalon, and ventricular enlargementwere visualised. Despite the clinical syn-drome and these secondary changes dueto compression, there was only approxi-mately 2 mm of downward displacementof the upper brainstem compared with 13mm horizontal displacement. Althoughtissue shifts adjacent to the tentorialaperture cause brainstem and vascularcompression, these changes may occurwith minimal downward herniation.

(JNeurolNeurosurg Psychiatry 1993;56:932-935)

In the presence of a large intracranial massthe true relationships of structures near thetentorial notch may not be appreciated aswell by postmortem examination as they areby radiological analysis during life. The CTand MRI features of herniation from a unilat-eral mass have been reported,l" and severalMRI scan measurements have been proposedto detect downward movement of the upperbrainstem,5-7 but rarely with clinical correla-tion. This paper demonstrates most of theknown primary and secondary radiologicalfeatures of transtentorial herniation by MRIscans from a comatose patient with a largeacute subdural haematoma. The relative ver-tical and horizontal displacements of theupper brainstem were measured in anadvanced stage of herniation to determine therelative contributions of each.

She had no oculocephalic eye movements. ACT scan showed a large left sided subduralhaematoma with 13 mm of horizontal shift ofthe pineal calcification from the midline, andthe perimesencephalic cisterns were obliter-ated. The MRI scan described below wasperformed 22 hours later, with the same signspresent. She died four days after admission.Necropsy was not obtained.

MRIfindingsAxial, coronal, and sagittal T-1 weightedimages were obtained. Axial scans just belowthe plane of the tentorium (fig 1) showedconvex distortion and extreme displacementof the upper brainstem away from the sub-dural clot, and medial temporal tissue in theperimesencephalic cistern on the side of themass. The upper pons showed less displace-ment but the trigeminal nerve on the side ofthe mass was stretched (fig 2). Coronalimages showed the distortion of the entire

St Elizabeth'sHospital, Boston, MA,USAA H RopperCorrespondence to:Dr Ropper, St Elizabeth'sHospital, 736 Cambridge StBoston, MA 02135, USA.Received 17 August 1992and in revised form25 January 1993Accepted 29 January 1993

Case reportAs 86 year old woman fell at home and wasfound unresponsive. On admission she wascomatose, the left pupil was 6 cm, and rightpupil 4 mm in diameter, both round andunreactive. The right side was flaccid andmotionless and the left side had extensor pos-turing. There were bilateral Babinski's signs.

Figure 1 Axial MRI, S mm thick, T-1 weighted, TR720/TE 15, showing the left uncus and parahippocampalgyrus within the perimesencephalic cistern. There is upperbrainstem displacement and distortion, and enlargement ofthe contralateral ventnicle.

932 on A

pril 11, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.56.8.932 on 1 August 1993. D

ownloaded from

Syndrome of transtentorial herniation: is vertical displacement necessary?

Figure 2 AxialMRI atthe level of the upper ponsshowing displacement ofthe brainstem awayfromthe mass, stretching of thetrigeminal nerve on the sideof the clot, and slackeningcontralaterally. The sixthnerves could not bevisualised.

Figure 3 Coronal T-1weighted images showingthe convex displacementawayfrom the clot andmedial temporal lobeherniation.

central brain region with temporal lobeencroachment on the cerebral peduncle (fig3). Left cingulate gyrus infarction was alsodemonstrated. Coronal scans showed thedeep tissue shifts best, including theencroachment of the left temporal lobe on theadjacent perimesencephalic cistern, but therewas a large degree of displacement of themidbrain-diencephalic region above this level,leaving unclear the role of the herniated tissuein causing brainstem displacement. Theoccipital ventricular horn contralateral to theclot was enlarged.T-2 weighted images showed infarction of

the lower diencephalon and upper midbrain,and infarctions in the territories of the poste-rior cerebral artery on the side of the clot, ofthe calcarine branch opposite the clot, and ofthe callosomarginal branch of the anteriorcerebral artery (fig 4).

Horizontal and vertical measurementsMeasurements were made on printed MRI aspreviously described using electroniccalipers,6 and confirmed by direct computermeasurements on the MRI screen. There wasless than 10% difference in measurementsmade by the two techniques.

Sagittal images and vertical displacement: Ona mid-sagittal T-1 weighted MRI, thepontmesencephalic junction (PMJ) was 5 mmabove Twining's line in true perpendiculardistance (fig 5), therefore representingapproximately 2 mm downward movementcompared with normal subjects fromFeldman's series5 and our own normal sub-jects. The ratio of the vertical perpendiculardistance between Twining's line and PMJ tothe length of Twining's line, a measurementutilised by Feldman et al,5 was 0-042, midwaybetween the normal subjects and patientswith masses and higher than their awake

Figure 4 Axial T-2 weighted images, TR 30001TE 80 demonstrating infarctions in theupper brainstem and in the regions of the anterior and posterior cerebral arteries on the sideof the mass and a small area of medial occipital infarction opposite the mass.

Figure 5 Sagittal T-1 weighted image with (1)Twining's line and (2) the perpendicular distance to thepontomesencephalic junction of 5 mm indicated (2 mmbelow its normal position). The iter of the aqueduct cannotbe clearly seen because of displacement out of the plane ofthe image, but it is situated less than 5 mm below theincisural line used by Reich, et al'.

933

on April 11, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.8.932 on 1 A

ugust 1993. Dow

nloaded from

934

patients with masses. This ratio again repre-sents a true dimension of approximately 1-2mm downward displacement of the PMJ(length of Twining's line was 10 cm). Animproved reference measurement for the dis-placement of the posterior midbrain, intro-duced by Reich et al,7 uses the displacementof the iter (upper opening of the aqueduct)from its normal position on a line approxi-mating the plane of the incisura. This couldnot be measured with precision because theaqueduct was compressed and the iter wasdisplaced horizontally out of the mid-sagittalplane (as it is in most patients with masses).An approximate measurement for this verticaldisplacement was less than 5 mm downward.

Coronal images and horizontal displacement:On coronal images (fig 3) there was 13 mmhorizontal displacement of the lowest point ofthe third ventricle from the midline truedimension (the same as pineal displacementon CT), and 37-8 mm vertical distance fromthe pontomesencephalic junction (PMJ) tothe cerebral midline point, representing 1 2mm depression of the PMJ below its normalposition compared with controls in a previousstudy.6The three sagittal measurements and one

coronal measurement of vertical displacementtherefore converged to estimate approxi-mately 2 mm downward displacement of theupper brainstem structure compared to 13mm lateral displacement of structures in thatregion. The technically less precise measure-ment between the iter and the incisural linesuggested, at most, 5 mm downward dis-placement of the posterior brainstem.

DiscussionMRI can now be performed on acutely illpatients and may give more insight thannecropsy studies into the pathoanatomy ofherniation. Almost all of the features oftranstentorial herniation due to a large unilat-eral mass that are known from pathologystudies could be appreciated in the MRIscans in this case. Accompanying thesechanges was, however, extreme horizontal butlittle axial downward displacement of theupper brainstem near the plane of the tentor-ial incisura. Infarctions in both the posteriorcerebral artery territories,8-14 and colloso-marginal branch of the anterior cerebralartery" 1516 territory were detected with T-2weighted images. Extreme signal changesoccurred in the compressed diencephalic-mesencephalic junction,'7 including the radio-logical features of the Kernohan's notchphenomenon.'8 These changes in the upperbrainstem presumably reflected damage inthe reticular activating system that causedcoma.

In analysing the regional anatomy of struc-tures near the tentorium, the advanced stageof tissue displacement and preceding 22hours of coma must be considered.Transtentorial herniation of the medial tem-proal lobe had occurred by the time of theMRI scan but it is difficult to know if this

degree of tissue distortion was necessary forcoma and pupillary changes. There was alarge uncal herniation on the side of the massand the ipsilateral pupil was 6 mm in diame-ter, the opposite, 4 mm, but without uncalherniation on that side. These scans do notsettle the precise relationship of herniation toupper brainstem compression. The parahip-pocampal gyrus may not be the proximatecause of compression and displacement of theupper brainstem; rather it may move into theperimesencephalic cistern passively as theentire hemisphere is displaced horizontallyabove.619 Although there were points of conti-guity between the cerebral peduncle and thedisplaced medial temporal lobe, the incisuralcistern could still be visualised on coronalscans, suggesting that the horizontal brain-stem displacement resulted from tissue shiftswell above the tenorial plane, and enlarge-ment of the cistern may have been a passivephenomenon.

There was concordance between horizontalshift of the pineal on CT scan and of thelower third ventricular shift on MRI scan.Midline structures such as the pontomesen-cephalic junction and aqueduct are movedout of the mid-sagittal plane by large hemi-spheral masses, making most measurementsin these patients only approximations. Thenewer reference line introduced by Reich etall is an improvement from Twining's linebut compression and displacement of theaqueduct also limits its use when there ishorizontal displacement. These inaccuraciesare compounded by the optical problems incalibrating measurements of this small size.As discussed previously,6 it is difficult tocompare the physiological effects on upperbrainstem neurons of small downward dis-placement compared with larger horizontaldisplacements. Nevertheless, even in theadvanced stage of medial temporal lobe her-niation in our patient with coma, pupillarychanges, and bilateral compression of majorvessels at the tentorial aperture, there was lit-tle downward "herniation" of the upperbrainstem, contrasted with substantial hori-zontal displacement.

1 Osborn AG. Diagnosis of transtentorial herniation by cra-nial computed tomography. Radiology 1977;123:93-6.

2 Stovring J. Descending tentorial herniation: Findings oncomputed tomography. Neuroradiology 1977;14: 101-5.

3 Nguyen JP, Djindjian M, Brugieres P, Badiane S, MelonE, Poirer J. Correlations anatomo-scanographiquesdans les engagements cerebreaux transtentoriels. J7Neuroradiology 1989;16: 181-96.

4 Hahn FJ. Signs of central descending transtentorial hemi-ation. AJNR 1985;6:844-5.

5 Feldman E, Gandy SE, Becker R, Zimmerman R, ThalerHT, Posner JB, Plum F. MRI demonstrates descendingtranstentorial herniation. Neurology 1988;38:697-701.

6 Ropper AH. A preliminary MRI study of the geometry ofbrain displacement and level of consciousness withacute intracranial masses. Neurology 1989;39:622-7.

7 Reich JB, Sierra J, Deck MDF, Plum F. MRS descriptionand clinical correlation of dynamic upward and down-ward transtentorial herniation. Neurology 1991;41(suppl):390-1.

8 Sato M, Tanaka S, Kohama A, Fujii C. Occipital lobeinfarction caused by tentorial herniation. Neurosurgery1986;18:300-5.

9 Marinkovic SV, Milisavljevic MM, Lolic-Dragnic V,Kovacevic MS. Distribution of the occipital branches ofthe posterior artery. Correlation with occipital lobeinfarcts. Stroke 1987;18:728-32.

10 Blinkov SM, Gabibov GA, Tanyashin SV. Variations in

Ropper

on April 11, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.8.932 on 1 A

ugust 1993. Dow

nloaded from

Syndrome of transtentorial herniation: is vertical displacement necessary?

location of the arteries coursing between the brainstemand the free edge of the tentorium. Neurosurgery1992;76:973-8.

11 Lindenberg R. Compression of brain arteries as patho-genetic factor for tissue necroses and their areas ofpredilection. Jf Neuropathology Exp Neurology 1985;14:223-30.

12 Evans JP, Scheinker IM. Histologic studies of the brainfollowing head trauma. UI. Post traumatic infarction ofcerebral arteries, with consideration of the associatedclinical picture. Arch Neurology Psychiatry 1943;50:258-78.

13 Moore MT, Stern K. Vascular lesions in the brain-stemand occipital lobe occurring in association with braintumors. Brain 1938;61:70-98.

14 Sunderaland S. The tentorial notch and complicationsproduced by herniations of the brain through that aper-

ture. BritJSurgery 1958;45:422-38.15 Rothfus WE, Goldberg AL, Tabas JH, Deeb ZL.

Callosomarginal infarction secondary to transfalcial her-niation. AJNR 1987;8: 1073-6.

16 Sohn D, Levine S. Frontal lobe infarctions caused bybrain herniation. Compression of the anterior cerebralartery branches. Arch Pathology Lab Medicine 1967;84:509-12.

17 Niikawa S, Uno T, Ohkuma A, Hara A, Nokura H,Yamada H. Occlusion of a perforating artery indicatingdescending tentorial herniation after head injury, sup-plying deep cerebral structure-report of 4 cases andtheir CT evaluation. No To Shinkei 1988;40:1151-6.

18 Cohen AR, Wilson J. Magnetic resonance imaging ofKemohan's notch. Neurosurgery 1990;27:205-7.

19 Fisher CM. Acute brain herniation: a revised concept.Seminars in Neurology 1984;4:417-21.

935 on A

pril 11, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.56.8.932 on 1 August 1993. D

ownloaded from