0 evaluări0% au considerat acest document util (0 voturi)
202 vizualizări13 pagini
Cisterna magna septa, fourth ventricle, and cerebellar vallecula at various stages of development were studied. Septa are typically seen inferoposterior to the cerebellar vermis, usually straight and parallel. Findings support current opinion that persistent Blake's pouch and mega cisterna maga represent (less severe) abnormalities within the Dandy-Walker continuum.
Cisterna magna septa, fourth ventricle, and cerebellar vallecula at various stages of development were studied. Septa are typically seen inferoposterior to the cerebellar vermis, usually straight and parallel. Findings support current opinion that persistent Blake's pouch and mega cisterna maga represent (less severe) abnormalities within the Dandy-Walker continuum.
Cisterna magna septa, fourth ventricle, and cerebellar vallecula at various stages of development were studied. Septa are typically seen inferoposterior to the cerebellar vermis, usually straight and parallel. Findings support current opinion that persistent Blake's pouch and mega cisterna maga represent (less severe) abnormalities within the Dandy-Walker continuum.
Vestigial Remnants of Blakes Pouch and a Potential
New Marker for Normal Development of the Rhombencephalon Ashley J. Robinson, MB, ChB, Ruth Goldstein, MD Objective. The purpose of this study was to show the normal sonographic embryologic anatomy of the cisterna magna septa, fourth ventricle, and cerebellar vallecula at various stages of development and our experience with their variable appearance in multiple planes and to discuss the probable rela- tionship between the cisterna magna septa, Dandy-Walker continuum, mega cisterna magna, and per- sistent Blakes pouch. Methods. Retrospective and prospective selection of examples of cisterna magna septa was performed over approximately a 12-month period. Standard and nonstandard imag- ing planes were adopted as necessary. Results. The septa are typically seen inferoposterior to the cere- bellar vermis, usually straight and parallel, arising at the cerebellovermian angle and coursing posteriorly to the occipital bone. The cisterna magna septa become contiguous with the roof of the fourth ventricle inferior to the cerebellar vermis. The cerebrospinal fluid space enclosed between the cisterna magna septa is in direct contiguity with the fourth ventricle via the vallecula and is always com- pletely anechoic because it develops intra- and not extra-axially. Conclusions. We propose that the cis- terna magna septa represent the walls of Blakes pouch, a phylogenetic vestigial structure observed during ontogeny. Additionally, our observations support current opinion that a persistent Blakes pouch and mega cisterna magna represent (less severe) abnormalities within the Dandy-Walker continuum. The cisterna magna septa therefore are a marker of normal development of the roof of the rhomben- cephalon. Deviation from their normal appearances should prompt a closer assessment for associated abnormalities of the cerebellum, vermis, and brain stem by additional imaging in orthogonal planes with either sonography or magnetic resonance imaging. Key words: Blakes pouch; cisterna magna; Dandy-Walker continuum; mega cisterna magna; septa. Received June 7, 2006, from the Department of Radiology, Childrens Hospital of British Columbia, Vancouver, British Columbia, Canada (A.J.R.); and Department of Radiology, University of California, San Francisco, San Francisco, California USA (R.G.). Revision requested June 28, 2006. Revised manuscript accepted for publication August 24, 2006. Address correspondence to Ashley J. Robinson, MB, ChB, Department of Radiology, Childrens Hospital of British Columbia, 4480 Oak St, Vancouver BC V6H 3V4, Canada. E-mail: ash@radiologist.net Abbreviations CSF, cerebrospinal fluid; MRI, magnetic resonance imaging inear echoes are often seen in the fetal and neonatal cisterna magna by sonography. These septa have been variously described as the straight sinus, 1 the torular Herophili, 2 dural folds of the inferior attachments of the falx cerebelli, 3 or bridg- ing arachnoid septations. 4 Clearly there is disagreement about what they represent. Even in previous studies, which attempted pathologic correlation between the sonographic appearances and the gross and histologic appearances in fetal specimens, 3,4 there was disagree- ment in what these septations appeared to represent, one finding dense connective tissue containing small blood vessels but no arachnoid cells thought to be consistent with dura and the other finding concentrations of arach- noid septations; therefore, it seems unlikely that further pathologic studies will be of assistance. It is, however, rec- 2007 by the American Institute of Ultrasound in Medicine J Ultrasound Med 2007; 26:8395 0278-4297/07/$3.50 L Image Presentation 26.1.jum.online.q 12/12/06 9:10 AM Page 83 ognized that there is great difficulty in seeing membranes within the posterior fossa at autopsy unless the specimen is examined under water 5 and also because of specimen degeneration due to autolysis. 6 Characteristics of the Cisterna Magna Septa Previous studies have shown that the cisterna magna septa are seen in 84% to 92% of fetuses in the second and third trimesters. 3,4 The fact that they are seen in such a high number of fetuses is probably because they are normal structures. 4 Their visualization is not dependent on gesta- tional age after 15 weeks 3 and is dependent on the size of the cisterna magna, being most often observed when the cisterna magna is greater than 3 mm in anteroposterior diameter. 3 If they are not seen, it is often because of technical or positional issues or skull ossification. 4 These same previous studies showed that the cisterna magna septa are typically seen infero- posterior to the cerebellar vermis. They are usu- ally straight and parallel to the anteroposterior diameter of the cisterna magna 3 arising at the cerebellovermian junction 4 and coursing posteri- orly to the occipital bone (Figure 1). Serial follow- up imaging of the septa show them to be more difficult to discern later in gestation, and they may even disappear altogether, but the posterior fossa structures can be seen to develop normally (Figure 2). Additional observations not previously described are that the cisterna magna septa become con- tiguous with the roof of the fourth ventricle inferi- or to the cerebellar vermis, as shown on fetal (Figure 3) and neonatal (Figure 4) sonography. The cerebrospinal fluid (CSF) space enclosed between the cisterna magna septa is therefore in direct con- tiguity with the CSF space in the fourth ventricle, via the vallecula. Also, the fluid between the septa is always completely anechoic, whereas the fluid outside the septa is always slightly echogenic (Figures 13 and 5), although this effect is reduced throughout gestation. Often, an additional midline septum is visible in the posterior fossa on fetal and neonatal sonography, which represents the falx cerebelli (Figure 5). Therefore, 1, 2, or 3 septa can be observed depending on the scan plane: some- times only the midline septum (falx cerebelli) straight to the occiput, sometimes only the 2 par- allel septa originating at the cerebellovermian junction, and sometimes all 3 septa. 4 This triple- septa appearance relates not only to the scan plane but also to the apparent variable inferior extent of the falx cerebelli. 84 J Ultrasound Med 2007; 26:8395 The Cisterna Magna Septa Figure 1. Typical appearance of the cisterna magna septa (arrows) at 20 weeks gestation. Cb indicates cerebellum; and Cp, choroid plexus in the lateral ventricle. A, Semiaxial view. B, Coronal view. A B 26.1.jum.online.q 12/12/06 9:10 AM Page 84 Embryology The Dura At every site in the brain where the dural leaves are separated, a venous sinus exists in the space between them. This is because embryologically the sinuses develop from rich vascular plexuses within the epidural conjunctive tissue (Figure 6). 7,8 Therefore, the cisterna magna septa cannot represent dural leaves because we would expect there to be vascular spaces between them, but no vascular flow is ever shown. 3,4 Although there J Ultrasound Med 2007; 26:8395 85 Robinson and Goldstein Figure 3. Typical appearances of the cisterna magna septa (large arrows) at 13 weeks gestation. There was normal follow-up at 18 weeks. A, Above the level of the vallecula, the fourth ventricle (asterisk) is separated from the septa by developing cerebellar tonsils (small arrows). B, Slightly more inferiorly, the walls of the fourth ventricle are contiguous with the cisterna magna septa via the val- lecula (arrowhead). A B Figure 2. Typical appearance of the cisterna magna septa (arrows). Cb indicates cerebellum. A, At 22 weeks. B, Same fetus with nor- mal follow-up at 35 weeks. A B 26.1.jum.online.q 12/12/06 9:10 AM Page 85 are dural folds in the region of the cisterna magna, they cover tributary veins of the inferior sagittal sinus. 9 Additionally, they meet in the midline at the falx cerebelli and diverge around the foramen magnum, whereas the cisterna magna septa are parallel and do no meet in the midline. For similar reasons, the septa do not represent the walls of the straight sinus, torcular Herophili, or any other vascular structure. Also, the vein of Galen and straight sinus are located above the cisterna magna. 10 The Roof of the Rhombencephalon Early in development, there is focal dilatation of the central canal of the neural tube in the hind- brain. This can be shown by sonography as a cystic structure at the dorsal aspect of the fetal head (Figure 7). This is the rhombencephalic vesicle, and it eventually develops into the fourth ventricle. Normal development of the cerebellar vermis in the superior roof of the rhombencephalic vesicle has been shown on fetal sonographic 11,12 and specimen magnetic resonance imaging (MRI) studies, 1316 and, allowing for some physiologic variation in tim- ing, a closed (covered) fourth ventricle should usually be present by 17.5 to 18 weeks 17 but may be as late as 22 to 24 weeks. 11,12 The inferior roof of the rhombencephalic vesi- cle evaginates caudal to the developing cerebel- lar vermis into the meninx primitiva as an ependyma-lined diverticulum, 18 which expands 86 J Ultrasound Med 2007; 26:8395 The Cisterna Magna Septa Figure 4. Typical appearances of the cisterna magna septa (arrows) at 28 weeks on neonatal head sonography. Asterisk indicates fourth ventricle; and Cb, cerebellum. A, Above the vallecula, the fourth ventricle and septa (large arrows) are separated by the inferior vermis (small arrow). B, At the level of the vallecula, the septa (arrows) are contiguous with the walls of the fourth ventricle (arrowheads). A B Figure 5. Triple appearance of the cisterna magna septa on neonatal head sonography at 27 weeks gestation. The falx cere- belli is separate from the midline cerebellar structures. It becomes progressively smaller in anteroposterior diameter toward the infe- rior aspect of the posterior fossa until it disappears, owing to its sickle shape. The visible downside cisterna magna septum meets the brain at the cerebellovermian junction. Arrow indicates falx cerebelli; Cb, cerebellum; and Vm, cerebellar vermis. 26.1.jum.online.q 12/12/06 9:10 AM Page 86 dorsally to form Blakes pouch. 19 The walls of Blakes pouch are composed of an inner layer of ependyma, an intermediate layer of attenuated neuroglial tissue, and an outer layer of pia-arach- noid. 18 Blakes pouch fenestrates to a variable degree down to the obex (the inferior recess of the fourth ventricle), depending on whether the walls of the pouch are supported from adjacent neural and pial structures. 19,20 This fenestration leaves Blakes metapore. 20,21 The metapore, otherwise known as the foramen of Magendie, represents the neck of Blakes pouch, and the cisterna magna septa represent the remnants of its walls. This explains why the fluid between the septa is always anechoic, whereas the fluid lateral to the septa is usually slightly echogenic; the fluid between the septa is intra-axial, whereas the fluid outside the septa is in the subarach- noid space of the cisterna magna. The sub- arachnoid space is trabeculated in multiple planes by pia-arachnoid septations, a result of its embryologic development, 4 which give rise to the relative echogenicity compared with the intra-axial CSF spaces. The relative echogenic- ity diminishes throughout gestation as cavita- tion progresses. J Ultrasound Med 2007; 26:8395 87 Robinson and Goldstein Figure 6. Embryologic features of the dura. A, The brain (large arrow) is covered by a layer of meninx primitiva (small arrow) and vascular mesenchyme (double arrow). B, As the cerebral and cerebellar hemispheres expand and overgrow the developing midline structures (arrows), there is eventual apposition and fusion of the two reflected dural layers from adjacent hemi- spheres. The meninx primitiva starts to cavitate to form the sub- arachnoid space, delimited by an inner condensation of pia mater and an outer condensation of arachnoid mater. The vas- cular mesenchyme persists as a wedge of tissue trapped in the dural folds (double arrow). C, The reflected dural layers fuse, thus forming the falx cerebri, falx cerebelli, and tentorium. The meninx primitiva cavities coalesce to form the subarachnoid space, which is traversed by multiple pia-arachnoid septations (small arrow). Where no dural fusion takes place, the vascular mesenchyme cavitates to form venous sinuses, for example, the superior and inferior sagittal sinuses (double arrows). D, Color Doppler examination at 21 weeks showing flow in the occipital sinus (double arrow) between the diverging dural leaves of the falx cerebelli but none between or within the cisterna magna septa (arrows). C D A B 26.1.jum.online.q 12/12/06 9:10 AM Page 87 The process of dorsal expansion also explains why the tela choroidea has been shown to extend outward beneath the vermis into the cisterna magna 19,20 as it is carried in the walls of Blakes pouch as it expands dorsally. Once fenestration of the pouch occurs, it would then appear as if the tela choroidea would be within the cisterna magna and contiguous with the pia mater over the brain stem. Lateral Displacement of the Cisterna Magna Septa The fenestration of Blakes pouch and therefore communication of the pouch with the sub- arachnoid space are variable in timing but should occur by 4 months. 18,20,21 The foramen of Luschka normally opens later than the fora- men of Magendie; however, in 1% to 2% of healthy individuals, the foramen of Magendie is absent. 20 This is understandable inasmuch as ontogeny recapitulates phylogeny, and in lower species, the pouch never fenestrates, and Blakes metapore is therefore absent. 19 In these healthy individuals (and in lower species), it is the opening of the foramen of Luschka that leads to equilibrium of CSF between ventricles and subarachnoid cisterns. 22 We speculate that before fenestration of the foramen, there is transient increased intra-axial CSF pressure, and this explains the enlargement of the fourth ventricle, which is seen around 14 to 16 weeks gestation, resolving completely by 22 to 23 weeks gestation. 12,18,22 It also probably explains the outward bowing of the cisterna magna septa often observed sonographically, giving the impression of a posterior fossa cyst (Figure 8). 4 Intraventricular Hemorrhage Sometimes the septa are bowed laterally by the presence of intraventricular hemorrhage extend- ing posterior to the vallecula (Figure 9). Blood in the cisterna magna has previously been shown to have a high positive predictive value for develop- ment of posthemorrhagic hydrocephalus. 23 The likely explanation is that in the premature neonatal population, blood seen inferoposterior to the cerebellar vermis is actually not in the cis- terna magna but is contained within Blakes pouch and is therefore intra-axial, not extra- axial. Therefore, it has the potential to cause hydrocephalus because its intra-axial location allows it to block the foramina of Luschka and Magendie. 88 J Ultrasound Med 2007; 26:8395 The Cisterna Magna Septa Figure 7 (opposite page). Embryologic features of the develop- ing hindbrain. A, Sagittal view of an embryo at approximately 7 weeks gestation showing the typical appearances of the rhombencephalic vesicle (arrow). The primitive dorsal pontine flexure is also clearly shown (dotted line). B, During formation of the dorsal pontine flexure of the rhombencephalon (small arrow), a transverse crease forms in the roof of the rhomben- cephalic vesicle (large arrow), sandwiching a layer of the meninx primitiva. The crease divides the roof of the fourth ventricle (asterisk) into the anterior (superiorly) and posterior (inferiorly) membranous areas. C, The transverse crease and its wedge of vascular meninx primitiva develop into the choroid plexus (large arrow) of the roof of the fourth ventricle. The anterior membra- nous area thickens and proliferates posteroinferiorly, giving rise to the cerebellar vermis (arrowhead), which covers the posterior membranous area. The meninx primitiva (double arrow) cavi- tates to form the subarachnoid space. D, As a result of growth of overlying and surrounding structures, the posterior mem- branous area is left effectively evaginating posteriorly into the meninx primitiva between the inferior vermis superiorly (large arrow) and the nucleus gracilis inferiorly (small arrow), forming Blakes pouch, and carries the developing choroid plexus in its roof. The cavitations in the meninx primitiva coales- cence to form the subarachnoid space trabeculated by multiple pia-arachnoid septations. The fluid contained by Blakes pouch is intra-axial, even though the pouch itself lies within the devel- oping cisterna magna (double arrow). The pia mater (gray out- line) develops as a condensation of the meninx primitiva overly- ing the developing neural structures. The ependyma lining the neural tube (black outline) also lines Blakes pouch. As the pouch expands, its walls and the associated tela choroidea become apposed to the medial surfaces of the adjacent cerebellar hemi- spheres and the undersurface of the superjacent inferior vermis. Because these also have pia mater over their outer surfaces, the two outer layers of pia mater become apposed and indistin- guishable. E, Blakes pouch fenestrates, leaving an opening at its neck called Blakes metapore (dashed oval). The choroid plexus (arrow) is now in direct contiguity with the pia mater over the inferior surface of the cerebellar vermis and lies in the (extra- axial) subarachnoid space, although it developed intra-axially. The remnants of Blakes pouch probably persist to a variable degree. F, In the axial plane, Blakes pouch extends dorsal to the fourth ventricle (asterisk) into the developing subarachnoid space below the vermis and between the cerebellar hemi- spheres. The developing choroid plexus (double arrows) in the roof of the rhombencephalic vesicle has normal caudal exten- sions on either side of the midline, which, once Blakes pouch fenestrates, will appear to lie on the extra-axial surface of the cerebellum. The lateral recesses of the rhombencephalic vesicle (small arrows) are also in contact with the developing subarach- noid space and fenestrate to form the foramen of Luschka. The walls of Blakes pouch (large arrows) are visible sonographically as the cisterna magna septa. 26.1.jum.online.q 12/12/06 9:10 AM Page 88 Mega Cisterna Magna A mega cisterna magna (cisterna magna 10 mm; Figure 10) is known to result from a defect of the posterior membranous area during embryogene- sis. 21 We speculate that in this condition, Blakes pouch probably persists and expands posterior to a normally formed cerebellum and vermis, but unlike a persistent Blakes pouch (see below), fen- estration occurs late, leaving behind an expanded posterior fossa, thus explaining the free communi- cation between the cyst, fourth ventricle, and subarachnoid space that is seen in this condition. 21 J Ultrasound Med 2007; 26:8395 89 Robinson and Goldstein F D A B E C 26.1.jum.online.q 12/12/06 9:10 AM Page 89 Dandy-Walker Continuum The observation that the cisterna magna septa were not seen in 16 of 18 fetuses with Dandy- Walker malformation 3 is most likely due to the fact that the cisterna magna septa (the walls of Blakes pouch) and the walls of the Dandy- Walker cyst are identical anatomic structures, and during expansion of the pouch within the cisterna magna, its walls become apposed to and indistinguishable from the sides of the pos- terior fossa (Figure 11). In those 2 of 18 cases in which the septa were seen, the size of the Dandy-Walker cyst may simply have been small- er, allowing them to be seen separately, 3 or the diagnosis was erroneous, perhaps representing a persistent Blakes pouch (see below). Persistent Blakes Pouch In a persistent Blakes pouch (Figure 12), there is thought to be failure of fenestration of both Blakes pouch and the foramen of Luschka, leading to dilatation of the fourth ventricle. 21 Although the pouch communicates freely with the fourth ventricle, there is a failure of commu- nication between the pouch and the per- imedullary subarachnoid spaces. 20,24 After ventricular shunting, subtotal or total reexpan- sion of the cerebellar hemispheres and vermis can be seen 22 ; that is, although these structures appear to be hypoplastic, there is actually no intrinsic hypoplasia but just the mass effect of the cyst. 90 J Ultrasound Med 2007; 26:8395 The Cisterna Magna Septa Figure 8. Septa appear bowed out (arrows), giving a cystic appearance in this neonate at 24 weeks gestation. Asterisk indi- cates fourth ventricle; and Cb, cerebellum. This appearance was seen in 18% of cases in a previous study. 4 Figure 9. Dilatation of the space between the cisterna magna septa on neonatal head sonography at 31 weeks gestation, secondary to intraventricular hemorrhage. A, Semiaxial view showing a dilated space between the cisterna magna septa with blood layering on the head-down (dependent) side septum (double arrow) but not reaching the subarachnoid space over the cerebellar hemisphere (small arrow). Asterisk indicates fourth ventricle; and Cb, cerebellum. B, Axial computed tomography (different patient) showing blood in the cisterna magna but on closer observation can be seen to be limited in its lateral extent by invisible walls (arrows) corresponding to the cisterna magna septa and indirectly indicating the existence of Blakes pouch, which itself cannot be resolved. A B 26.1.jum.online.q 12/12/06 9:10 AM Page 90 J Ultrasound Med 2007; 26:8395 91 Robinson and Goldstein Figure 10. Suspected mega cisterna magna at 19 weeks gestation. A, Axial scan below the level of the vallecula showing the cisterna magna septa (arrows) bowed outward. B, Axial scan at level of the vallecula showing the fourth ventricle (asterisk) to be contiguous with space between septa via the vallecula. Arrows indicate cerebellar ton- sils. C, Slightly oblique midline image showing evagination of the roof of the rhombencephalic vesicle forming a pouch (large arrow) con- taining anechoic fluid (CSF). The inferior border of the tentorium and falx cerebelli is visible (double arrow), with a slightly echogenic sub- arachnoid space directly inferior to it. The vermis (small arrow) is slightly compressed superoanteriorly. D, Sagittal MRI from the same fetus showing a mega cisterna magna (double arrow). The walls of the pouch are not detectable by MRI. The vermian size and morpho- logic characteristics are normal for gestation with a normal primary fissure (large arrow) and fastigial point (small arrow) but an impres- sion of slight compression inferiorly. E, In a mega cisterna magna, the foramina of Luschka and Magendie fenestrate late, leading to dilata- tion of the pouch and enlargement of the posterior fossa (double arrow). Once fenestration occurs, the vermis, which has developed normally but which may have been elevated, returns to a more nor- mal position. An enlarged subarachnoid space results, but actual- ly this enlarged space initially was intra-axial, not extra-axial. B D E A C 26.1.jum.online.q 12/12/06 9:10 AM Page 91 It has been shown that an increased angle between the vermis and tegmentum without any other abnormality of growth or structure of the vermis or abnormality elsewhere in the fetus may simply represent elevation of the ver- mis and does not necessarily indicate an adverse outcome, 5,2528 which could be explained by a persistent Blakes pouch. Unfortunately, this elevation of the vermis often leads to the false- positive diagnosis of a Dandy-Walker continu- um on axial sonography 25 and inferior vermian hypoplasia on sagittal sonography or MRI. This would explain why there has historically been a poor correlation between sonographic and autopsy findings in apparent cystic malforma- tions of the posterior fossa. 6 Dandy-Walker Continuum, Persistent Blakes Pouch, and Mega Cisterna Magna: A Single Spectrum We speculate that the spectrum of findings with respect to the posterior fossa cyst in the Dandy- Walker continuum, persistent Blakes pouch, and mega cisterna magna might depend on the degree of dilatation of Blakes pouch, which in turn depends on the relative timing of fenestra- tion of the fourth ventricle outlet foramina. It would appear that in a persistent Blakes pouch, the vermis is fully formed, but neither the pouch nor the foramina of Luschka fenestrate, leading to elevation of an otherwise normal ver- mis. In the Dandy-Walker continuum, there is 92 J Ultrasound Med 2007; 26:8395 The Cisterna Magna Septa Figure 11. Dandy-Walker continuum at 22 weeks gestation (case courtesy of Ants Toi, MD, and Susan Blaser, MD, Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada). The pregnancy was terminated. A, Axial view showing dilatation of space between the cisterna magna septa (arrows), which are effaced to adjacent structures and almost imperceptible. The fourth ventricle (asterisk) is contiguous with this space via the dilated vallecula. Cb indicates cerebellum; and Cp, cerebral peduncles. B, Sagittal MRI showing a hypoplastic cerebellar vermis (double arrow) and superior roof of the expanded Blakes pouch (small arrow). Adapted with permission from Imaging the Central Nervous System of the Fetus and Neonate. 27 C, In the Dandy-Walker continuum, there is variable vermian hypoplasia (small arrow), elevation by the dilated Blakes pouch (double arrow), and variable degrees of fenestra- tion of the foramina of Luschka and Magendie. A B C 26.1.jum.online.q 12/12/06 9:10 AM Page 92 J Ultrasound Med 2007; 26:8395 93 Robinson and Goldstein Figure 12. Persistent Blakes pouch at 23 weeks (case courtesy of Ants Toi, MD, and Susan Blaser, MD, Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada). The fetus was stillborn at 28 weeks because of placental insufficiency in a mother with sickle cell disease. The karyotype and pathologic findings were both normal. A, Axial view with an expanded fourth ven- tricle (asterisk) and neck of Blakes pouch causing a mass effect with concavity of the adjacent cerebellar hemispheres (Cb), which appear normal in size. Cp indicates cerebral peduncles. B, Sagittal MRI showing a cerebellar vermis (double arrow) with normal land- marks that has normal biometric characteristics, but it is angled away from the tegmentum (small arrow), with a large gap between the inferior vermis and brain stem (arrowhead). Because the vermian structure is normal, this gap should not be confused with infe- rior vermian hypoplasia. Adapted with permission from Imaging the Central Nervous System of the Fetus and Neonate. 27 C, Sagittal histopathologic specimen showing that the vermis has returned to a normal position as Blakes pouch has collapsed. The fourth ven- tricle is now normal in size (asterisk). The vermian structure is normal, with normal primary (long arrow) and prepyramidal (short arrow) fissures. It is important to note that in the fully mature vermis, there will be 3 lobules between these 2 landmarks (declive, folium, and tuber), but unlike the other vermian lobules, their white matter cores unite before reaching the fastigium (circle). On fetal MRI, this can lead to underestimation of the number of visible vermian lobules. Reproduced with permission from Imaging the Central Nervous System of the Fetus and Neonate. 27 D, In a persistent Blakes pouch, the cerebellar vermis forms normally. Nonfenestration of the foramina of Luschka and Magendie leads to dilatation of the fourth ventricle and Blakes pouch and elevation of the vermis (small arrow) away from the brain stem. There is no demonstrable communication between the cyst (double arrow) and the subarachnoid space by contrast cisternography. C D A B 26.1.jum.online.q 12/12/06 9:10 AM Page 93 variable vermian hypoplasia and elevation and variable fenestration. 29 In a mega cisterna magna, the vermis is fully formed, but fenestra- tion of Blakes pouch and the foramen of Luschka is delayed, thus allowing sufficient CSF equilibra- tion for the vermis not to be elevated but only after it has already caused expansion of the cis- terna magna. Conclusions We propose that the cisterna magna septa repre- sent the walls of Blakes pouch, a phylogenetic vestigial structure observed during ontogeny, the remnants of which are probably present in the neonatal period and possibly even into adult- hood. Additionally, our observations support current opinion that a persistent Blakes pouch and mega cisterna magna may represent (less severe) abnormalities within the Dandy-Walker continuum. The cisterna magna septa are therefore a potential new marker of normal development of the roof of the rhombencephalon. Deviation from their normal appearances should prompt a closer assessment for associated abnormalities of the cerebellum, vermis, and brain stem by additional imaging in orthogonal planes with either sonography or MRI. References 1. Mahony BS, Callen PW, Filly RA, Hoddick WK. The fetal cis- terna magna. Radiology 1984; 153:773776. 2. Pilu G, Romero R, De Palma L, Jeanty P, Burdine C, Hobbins JC. Ultrasound investigation of the posterior fossa in the fetus. Am J Perinatol 1987; 4:155159. 3. Pretorius DH, Kallman CE, Grafe MR, Budorick NE, Stamm ER. Linear echoes in the fetal cisterna magna. J Ultrasound Med 1992; 11:125128. 4. Knutzon RK, McGahan JP, Salamat MS, Brant WE. Fetal cis- terna magna septa: a normal anatomic finding. Radiology 1991; 180:799801. 5. Nelson MD Jr, Maher K, Gilles FH. A different approach to cysts of the posterior fossa. Pediatr Radiol 2004; 34: 720732. 6. Carroll SG, Porter H, Abdel-Fattah S, Kyle PM, Soothill PW. Correlation of prenatal ultrasound diagnosis and patholog- ic findings in fetal brain abnormalities. Ultrasound Obstet Gynecol 2000; 16:149153. 7. Lasjaunias P, Kwok R, Goh P, Yeong KY, Lim W, Chng SM. A developmental theory of the superior sagittal sinus(es) in craniopagus twins. Childs Nerv Syst 2004; 20:526537. 8. Muthukumar N, Palaniappan P. Tentorial venous sinuses: an anatomic study. Neurosurgery 1998; 42:363371. 9. Gray H. The meninges of the brain and medulla spinalis. In: Anatomy of the Human Body. 20th ed. New York, NY: Bartleby.com; 2000. Available at: http://www.bartleby.com/ 107/193.html. 10. Pilu G, De Palma L, Romero R, Bovicelli L, Hobbins JC. The fetal subarachnoid cisterns: an ultrasound study with report of a case of congenital communicating hydro- cephalus. J Ultrasound Med 1986; 5:365372. 11. Ben-Amin M, Perlitz Y, Peleg D. Transvaginal sonographic appearance of the cerebellar vermis at 1416 weeks ges- tation. Ultrasound Obstet Gynecol 2002; 19:208209. 12. Bronshtein M, Zimmer EZ, Blazer S. Isolated large fourth ventricle in early pregnancy: a possible benign transient phenomenon. Prenat Diagn 1998; 18:9971000. 13. Hansen PE, Ballesteros MC, Soila K, Garcia L, Howard JM. MR imaging of the developing human brain, part 1: pre- natal development. Radiographics 1993; 13:2136. 14. Nakayama T, Yamada R. MR imaging of posterior fossa structures of human embryos and fetuses. Radiat Med 1999; 17:105114. 15. Chong BW, Babcook CJ, Pang D, Ellis WG. A magnetic res- onance template for normal cerebellar development in the human fetus. Neurosurgery 1997; 41:924929. 16. Babcook CJ, Chong BW, Salamat MS, Ellis WG, Goldstein RB. Sonographic anatomy of the developing cerebellum: normal embryology can resemble pathology. AJR Am J Roentgenol 1996; 166:427433. 17. Bromley B, Nadel AS, Pauker S, Estroff JA, Benacerraf BR. Closure of the cerebellar vermis: evaluation with second trimester US. Radiology 194; 193:761763. 18. Strand RD, Barnes PD, Poussaint TY, Estroff JA, Burrows PE. Cystic retrocerebellar malformations: unification of the Dandy-Walker complex and the Blakes pouch. Pediatr Radiol 1993; 23:258260. 19. Blake JA. The roof and lateral recesses of the fourth ventri- cle considered morphologically and embryologically. J Comp Neurol 1900; 10:79108. 20. Raybaud C. Cystic malformations of the posterior fossa: abnormalities associated with the development of the roof of the fourth ventricle and adjacent meningeal structures. J Neuroradiol 1982; 9:103133. 21. Tortori-Donati P, Fondelli MP, Rossi A, Carini S. Cystic mal- formations of the posterior cranial fossa originating from a defect of the posterior membranous area. Mega cisterna magna and persisting Blakes pouch: two separate entities. Childs Nerv Syst 1996; 12:303308. 22. Calabro F, Arcuri T, Jinkins JR. Blakes pouch: an entity with- in the Dandy-Walker continuum. Neuroradiology 2000; 42:290295. 23. Cramer BC, Walsh EA. Cisterna magna clot and subse- quent post-hemorrhagic hydrocephalus. Pediatr Radiol 2001; 31:153159. 94 J Ultrasound Med 2007; 26:8395 The Cisterna Magna Septa 26.1.jum.online.q 12/12/06 9:10 AM Page 94 24. Arai H, Sato K. Posterior fossa cysts: clinical, neuroradio- logical and surgical features. Childs Nerv Syst 1991; 7:156 164. 25. Barkovich AJ, Kjos BO, Norman D, Edwards MS. Revised classification of posterior fossa cysts and cystlike malfor- mations based on the results of multiplanar MR imaging. AJR Am J Roentgenol 1989; 153:12891300. 26. Klein O, Pierre-Kahn A, Boddaert N, Parisot D, Brunelle F. Dandy-Walker malformation: prenatal diagnosis and prog- nosis. Childs Nerv Syst 2003; 19:484489. 27. Robinson AJ, Blaser S. In-utero MR imaging of develop- mental abnormalities of the fetal cerebellar vermis. In: Griffiths PD, Paley MN, Whitby EH (eds). Imaging the Central Nervous System of the Fetus and Neonate. Oxford, England: Taylor & Francis; 2006:115131. 28. Zalel Y, Gilboa Y, Gabis L, Ben-Sira L, Hoffman C, Wiener Y, Achiron R. Rotation of the vermis as a cause of enlarged cisterna magna on prenatal imaging. Ultrasound Obstet Gynecol 2006; 27:490493. 29. Hirsch JF, Pierre-Kahn A, Renier D, Sainte-Rose C, Hoppe- Hirsch E. The Dandy-Walker malformation: a review of 40 cases. J Neurosurg 1984; 61:515522. J Ultrasound Med 2007; 26:8395 95 Robinson and Goldstein 26.1.jum.online.q 12/12/06 9:10 AM Page 95