Cranial Meningocele

CSF Leak Associated with Persistent or Intermittent Rhinorrhea

When persistent or intermittent watery rhinorrhea suggests CSF leak, endoscopy is the first line examination. An endonasal expansile lesion (cephalocele? glioma?) may be detected or endoscopy may give a negative result. In the first case, BTP test is unnecessary, as CSF leak is assumed to be related to the endonasal lesion. Additionally, having detected a nasal mass, the main clinical issue is to defining its relationship with intracranial structures. In this setting, MR is preferable to high resolution CT. In fact, T2 and MR cisternographic sequences are superior in demonstrating the uninterrupted CSF signal extending from the subarachnoid space into the mass, or into paranasal sinuses/nasal cavity or into middle ear, with or without brain tissue (STaFFORD et al. 1996) (Fig. 7.6).

Meningocele Cranial Intranasal

Fig. 7.6a,b. Small spontaneous meningocele through the right cribriform plate. Coronal TSE T2 before (a) and after (b) surgical repair. a A small rounded fluid collection (proved to be a meningocele at surgery) is seen under the right cribriform plate (asterisk); both olfactory bulbs are clearly visible over the cribriform plate in this patient with a "deep" olfactory fossa (the left one indicated by arrows). Subtle mucosal thickening in both maxillary sinuses, with retention cyst on the right side. b After successful endoscopic repair with overlay technique, the CSF collection in the right olfactory cleft is no longer visible. Note the post-surgical changes on the right side due to middle turbinectomy, uncinectomy, middle antrostomy, and ethmoidotomy

Fig. 7.6a,b. Small spontaneous meningocele through the right cribriform plate. Coronal TSE T2 before (a) and after (b) surgical repair. a A small rounded fluid collection (proved to be a meningocele at surgery) is seen under the right cribriform plate (asterisk); both olfactory bulbs are clearly visible over the cribriform plate in this patient with a "deep" olfactory fossa (the left one indicated by arrows). Subtle mucosal thickening in both maxillary sinuses, with retention cyst on the right side. b After successful endoscopic repair with overlay technique, the CSF collection in the right olfactory cleft is no longer visible. Note the post-surgical changes on the right side due to middle turbinectomy, uncinectomy, middle antrostomy, and ethmoidotomy

CT has a complementary role, related to its superior depiction of bony detail. At MR, the cephalocele appears as a round "mass" with sharp margins, isointense to CSF in all sequences (meningocele) or containing some tissue isointense to brain parenchyma (meningoencephalocele), protruding from the in-tracranial cavity into the nasal cavity or paranasal sinuses. The lumen of the lesion is continuous with the subarachnoid space and often has a constriction in the portion passing through the bone defect (the "neck" of the cephalocele) (Fig. 7.7). CT clearly shows the sclerotic margins of the bone defect. Cerebral MR may show the associated brain anatomy "distortion", presumably due to the effect of the brain pulsation in utero, which "pushed" the pliable unmyelinated brain outward through the defect (Truwit et al. 1996). The result is a general tendency of the ventricles and subarachnoid spaces, which subtend the cephalocele, to be stretched and elongated, "pointing" toward the calvarial defect. This finding is usually observed in patients with large occipital or parietal cephaloceles.

Cephaloceles may be isolated anomalies, but may also be seen in conjunction with other congenital brain malformations or as a part of a syndrome. The radiologist must therefore also look for associated anomalies, such as agenesis or hypogenesis of the corpus callosum, Dandy-Walker malformation, or malformations of cortical development (Naidich et al. 1992).

Among lesions to be differentiated from cephalo-celes is the so-called nasal glioma (heterotopic brain tissue). It consists of a variable amount of dysplastic brain tissue (especially glia cells, whereas neurons are present in only about 10% of cases), located within the nasal cavity or in the nasal subcutaneous tissue. It is thought to result from the herniation of brain tissue into a dural projection that normally extends through the foramen cecum during the embryologic development. Regression of the more superior portion of this dura projection leads to complete separation of the nasal glioma from the intracranial contents. If infe-riorly the dural projection remains adherent to the skin of the nose and fails to involute, but no brain tissue herniates through it; a dermal sinus tract is present. It is usually suggested by a small dimple on the surface of the nose. Along the path of the dermal sinus tract (epi)dermoids may develop.

MR is clearly superior to CT in the evaluation of nasal gliomas, because it directly shows the lack of communication with the subarachnoid space, which is the hallmark of cephaloceles. Furthermore, the dysplastic brain tissue of nasal glioma appears hyper-

Intranasal Gliomas

Fig. 7.7a-c. Fronto-ethmoidal spontaneous meningocele in an 8-year-old girl. Coronal high resolution CT (a), TSE T2 in the coronal (b) and sagittal (c) planes. a Bone defect in the left cribriform plate (arrow), associated with soft-tissue density in the left nasal cavity (asterisk). b The outpouching of CSF in the left nasal cavity is clearly identified as fluid signal (hy-perintense) on TSE T2 images (asterisk). Even if the signal intensity of CSF is slightly brighter than that of nasal mucosa, an inflammatory polyp could be a differential diagnosis. c On sagittal image, the direct communication between the fluid collection in the nasal cavity and the frontal subarachnoid space is clearly depicted (arrow). There is no brain parenchyma extending into the endonasal CSF collection c

Meningozele Ethmoidal

Fig. 7.7a-c. Fronto-ethmoidal spontaneous meningocele in an 8-year-old girl. Coronal high resolution CT (a), TSE T2 in the coronal (b) and sagittal (c) planes. a Bone defect in the left cribriform plate (arrow), associated with soft-tissue density in the left nasal cavity (asterisk). b The outpouching of CSF in the left nasal cavity is clearly identified as fluid signal (hy-perintense) on TSE T2 images (asterisk). Even if the signal intensity of CSF is slightly brighter than that of nasal mucosa, an inflammatory polyp could be a differential diagnosis. c On sagittal image, the direct communication between the fluid collection in the nasal cavity and the frontal subarachnoid space is clearly depicted (arrow). There is no brain parenchyma extending into the endonasal CSF collection b a c intense to gray matter in both T1- and T2-weighted images (Fig. 7.8).

When no endonasal mass is detected by endos-copy, BTP test is indicated, its sensitivity for CSF leak being more than 90% (BAchmANN et al. 2000).

If post-traumatic rhinorrhea is investigated, positivity of BTP will confirm the diagnosis. In this setting, high resolution CT is the first line imaging technique, having a reported sensitivity ranging from 50% up to 100% (DiETrich et al. 1993; LLOYd et al. 1994, Zapalac 2002). It is obtained without intra-thecal or intravenous contrast agent administration, using thin (1- or 2-mm) contiguous axial and coronal planes. Scans or multiplanar reconstructions are orientated parallel and perpendicular to the anterior cranial fossa floor, respectively. The whole skull base needs to be examined, including the mastoid cells.

Because CSF leak is not directly detectable on high-resolution CT, the identification of a bony defect is indirectly taken as the possible site of the CSF leak, even if the bone defect itself does not necessarily correspond to the site of dural tear (Tolley et al. 1992). However, bony dehiscences are not infrequent in the skull base (Tolley et al. 1991), having been demonstrated in about 14% of ethmoid bones (Ohnishi 1981). Similar dehiscences have been described in the sphenoid and, less commonly, in the frontal sinus. As a consequence, a simple bony defect cannot be considered a reliable sign of CSF fistula. Conversely, when a bony defect, located at the edge between skull base and paranasal sinus/nasal cavity, is associated with fluid collection and/or mucosal thickening within the adjacent sinus/nasal cavity, it can be assumed to be located close to the dural breach site (Fig. 7.9). By combining these findings, the specificity of high-resolution CT raised to 86% in a series of 15 patients with traumatic (accident or iatrogenic) fistulae (Lloyd et al. 1994).

However, precise location of the fistula(e) may be rather difficult, if not impossible, in the presence of comminuted fractures, particularly when scar tissue partially replaces the skull base or invests its endona-sal interrupted surface. In this case, CT findings may indicate more than a single potential site of CSF leak, therefore not enabling a sufficiently tailored surgical approach. MR may provide additional findings, like showing signal intensity consistent with CSF within the scar tissue or across the interrupted skull base.

Nasal Glioma

Fig. 7.8a-e. Intranasal heterotopic brain tissue (nasal glioma). Coronal (a-c) and sagittal (d-e) TSE T2. a The mass (asterisk) arises from left olfactory fissure, the left nasal cavity is completely occupied. b The nasal glioma abuts the inferior surface of the cribriform plate (1) without any connection with subarachnoid space. Middle (2) and inferior (3) turbinates are compressed. c The vertical lamina of the middle turbinate limits the lateral aspect of the nasal glioma. d On midline sagittal plane, the crista galli (black arrows) shows intermediate signal. e On off-midline sagittal plane, the nasal glioma (asterisk) appears bordered superiorly by the lateral aspect of the crista galli (black arrows). A normal nasofrontal suture is present (white arrow)

Crista Galli

Fig. 7.8a-e. Intranasal heterotopic brain tissue (nasal glioma). Coronal (a-c) and sagittal (d-e) TSE T2. a The mass (asterisk) arises from left olfactory fissure, the left nasal cavity is completely occupied. b The nasal glioma abuts the inferior surface of the cribriform plate (1) without any connection with subarachnoid space. Middle (2) and inferior (3) turbinates are compressed. c The vertical lamina of the middle turbinate limits the lateral aspect of the nasal glioma. d On midline sagittal plane, the crista galli (black arrows) shows intermediate signal. e On off-midline sagittal plane, the nasal glioma (asterisk) appears bordered superiorly by the lateral aspect of the crista galli (black arrows). A normal nasofrontal suture is present (white arrow)

Retentioncyst Sinus

Figure 7.9. Traumatic iatrogenic CSF fistula, post-microendo-scopic sinus surgery. Coronal high resolution CT at the level of the olfactory fossa. A large bone defect is demonstrated to involve the horizontal (short arrows) and the lateral lamella (arrowhead) of the right cribriform plate. The right lamina papyracea is not recognizable (long arrows). Note the absence of the middle turbinate and ethmoid labyrinth on the right side, due to previous endonasal surgery

Figure 7.9. Traumatic iatrogenic CSF fistula, post-microendo-scopic sinus surgery. Coronal high resolution CT at the level of the olfactory fossa. A large bone defect is demonstrated to involve the horizontal (short arrows) and the lateral lamella (arrowhead) of the right cribriform plate. The right lamina papyracea is not recognizable (long arrows). Note the absence of the middle turbinate and ethmoid labyrinth on the right side, due to previous endonasal surgery

In particular, MR cisternography has been proposed as a non-invasive cisternography technique. It consists of turbo spin echo sequences with fat-suppression (to null signal from bone marrow) that greatly increase CSF signal intensity and suppress the background, providing heavily T2 images with detailed anatomic information of the subarachnoid cisterns. If necessary, a small number of images (usually between two and eight) may be compressed into composite images using maximum intensity projection algorithms. Composite images may be useful to detect fistulae with irregular and tortuous tracks (El Gammal and Brooks. 1994). In detecting CSF fis-tulae, MR cisternography yielded sensitivity of 87%, specificity ranging from 57% up to 100%, and accuracy of 78%-89% (El Gammal et al. 1998, Shetty et al. 1998). Although MR cisternography may be added to the diagnostic work-up, at present precise identification of the fistula(e) is made possible at surgery by fluorescein leakage detection, which is more accurate than imaging techniques.

In fact, even fistulae due to incomplete damage of the meninges causing permeation rather than a complete breech are shown with this technique. The fluo-

roscein tracer is administered via lumbar puncture, and a special light filter is necessary to endoscopically detect small fistulae. In case of spontaneous fistula, a positive BTP test requires the patient to be investigated either by CT, adding the high resolution CT technique to the standard examination of the head, and by MR. Detection of paranasal sinuses or temporal bone hyperpneumatization will prompt the accurate assessment of the whole skull base to rule out bone dehiscences. The presence of an empty sella with or without a fluid-like broad-based lesion on sphenoid walls may arouse the suspicion of benign intracranial hypertension. In this setting, intermittent CSF leak may be associated with other imaging findings suggesting idiopathic in-tracranial hypertension, as increased CSF amount surrounding the optic nerves along their intraorbital course (Suzuki et al. 2001). Furthermore, any fluid-like broad-based lesion hanging on the ex-tracranial surface of the skull base may indicate a cephalocele located beyond the limit of diagnostic endoscopy (Stone et al. 1999).

However, the presence of a high signal on T2 images within paranasal sinuses or mastoid cells or adjacent to the skull base may be due to thickened mucosa, mastoiditis or rhinosinusitis, and not necessarily to CSF accumulation. Therefore, a CSF fistula may be suspected whenever the high signal of the fistulous tract appears to be in direct continuity with the intracranial subarachnoid space. This point is relevant, as high signal intensities on T2 sequences are shown in up to 25% of patients examined by MR for non-sinonasal diseases (Moser et al. 1991).

Indirect evidence of CSF fistula is provided by a low-lying gyrus rectus when the leak is located at the cribriform plate area ("gyrus rectus sign"), probably related to the negative pressure created by the fistula (Shetty et al. 1998).

Finally, unexpected intracranial or extracranial tumors, hydrocephalus, or skull base inflammatory lesions may be demonstrated by CT or MR.

When BTP test, high resolution CT, and MR are all negative, the diagnosis of CSF has to be considered unlikely or spontaneous healing of the dural defect(s) might have occurred during the diagnostic work-up.

In a series of 42 spontaneous and post-traumatic CSF leaks examined by high-resolution CT, radionu-clide cisternography, and CT cisternography, spontaneous resolution of CSF leakage within 1 month from imaging studies was observed in all patients negative at high resolution CT (29%) (Stone et al. 1999).

How To Reduce Acne Scarring

How To Reduce Acne Scarring

Acne is a name that is famous in its own right, but for all of the wrong reasons. Most teenagers know, and dread, the very word, as it so prevalently wrecks havoc on their faces throughout their adolescent years.

Get My Free Ebook


Responses

  • Aaron
    What is a meningocele of the cribriform plate?
    3 years ago

Post a comment