Online newspaper of professor Yasser Metwally

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 29, 2009 —  Blood supply of the vertebrae follows the embryological pattern, branches from each segmental intercostal artery or lumbar artery supply the adjacent halves of two vertebrae, the lower half of one vertebra above, and the upper half of the one below and intervening disc region. This occurs because the adjacent parts of any tow vertebrae and intervertebral disc develop from the same somites. each vertebral body has developed from four somites, thus each vertebral body gets blood supply from four arterial systems. Inside the vertebral body the arterioles terminate as a tortuous loops under the epiphyseal endplates, where they behave functionally as end arteries. If these end arteries are blocked, an infarct may develop. Spread of infection through the arterial route offers an explanation for the frequent early localization of spinal infection in the very vascular juxta-epiphyseal, paradiscal areas of the vertebral body. In addition to the spread of infection through the arterial flow, it is possible that the epidural plexus of veins play a part in the localization of the infection in the epidural areas.

The epidural veins drain into Batson’s perivertebral plexus of veins. This plexus has ramifications into the base of the brain and chest wall and has free communication with the intercostal, lumbar and pelvic veins. Blood in the Batson plexus probably flows in all directions depending upon the movement of the chest, coughing, and straining. Retrograde flow of blood from infected viscera to the spine may be responsible for spread of infection from pelvic or abdominal organs to the spinal epidural space, and paraspinal area.

• Pathogenesis and Pathology:

Following the infection marked hyperemia and sever osteoporosis take place. Vertebral destruction takes place by lysis of vertebral bones which are softened, easily yield, and leads to compression. Necrosis lakes place also due to ischemic infarction of bones secondary to arterial occlusion due to thrombo-embolic phenomenon, endarteritis, and periarteritis.

Extension of the inflammatory process to the epidural space may occur either by direct extension of a local infective focus such as vertebral osteomyelitis, which is very common in tuberculous infection, or as a haemtogenous infection. Staphylococcal infection is usually of metastatic origin due to haematogenous spread from a distant lesion. This results in the formation of acute pyogenic epidural abscess. Acute pyogenic epidural abscesses are most frequently found in the dorsal spine and usually on the back of the cord, as the retromedullary epidural space, occupied by fat, is largest and very rich in venous plexus in this region. In acute pyogenic infection of the epidural space significant cord compression can occur before any infective bone destruction has time to develop.

The intervertebral disc space is not involved primarily, because it is a relatively avascular structure. The early involvement of the paradiscal regions of the vertebrae jeopardizes the nutrition of the disc, which is later invaded by the inflammatory process. Once the disc is invaded, destruction progress rapidly. Extension of the inflammatory process to the paravertebral area usually results in the formation of a paraspinal abscess.

Figure 1. Paradiscal vertebral column infection

In tuberculosis of the spine involvement of the vertebrae occur in 3 different ways.

• Typical paradiscal lesion

In this variety spread of the infection occurs by the way of arteries. Disc involvement occurs earlier and before the appearance of destructive changes in the vertebrae.

• Anterior type of involvement

This variety is due to extension of an abscess beneath the anterior longitudinal ligament and the periosteum. The infection may spread up and down stripping the anterior and posterior ligament and the periosteum from the front and the sides of the vertebral bodies. This results in loss of the periosteal blood supply and destruction of the anterolateral surface of contiguous vertebral bodies. Collapse of vertebral bodies and diminution of disc space is usually minimal and occur late.

• Central type of lesion

The central disease arises as a result of infection which starts from the center of the vertebral body, the infection probably reaches the center through Batson venous plexus. Diminution of disc space is minimal and paravertebral abscess is usually not marked. The vertebral body may initially expand like a tumour but towards the later stages, the diseased vertebra may show a concentric collapse.

In contrast to tuberculosis of the spine, other forms of infection such as acute epidural abscess commonly occurs in the absence of x ray evidence of osteomylitis. The destructive changes seen in TB., are not seen in acute pyogenic infection of the extradural space and vertebral column.

• Pathology of inflammatory involvement of the spinal cord

             1-Inflammatory edema

Edema of the spinal cord due to vascular stasis and/or toxins from the infected foci in the neighboring vertebrae is considered to be the cause of early neurological deficits.

              2-Infarction of the spinal cord

This is an important cause frequently underestimated, infarction is caused by end arteritis, periarteritis, or thrombosis of an important tributary to the anterior spinal artery caused by inflammatory reaction.

             3- Extra-dural mass

An abscess in the extra-dural space, composed of fluid pus, caseous material, or granulation tissue, can cause significant compression of the spinal cord. Sequestra from avascular portions of the diseased vertebral bodies or intervertebral disc may be responsible for narrowing of the spinal canal and pressure on the cord.

              4- Meningeal changes

This is common in tuberculous infection. Thick layer of tuberculous granulation tissue may contract and undergo cicatrisation in long standing cases, this epidural fibrosis may be responsible for some cases of recurrent paraplegia.

              5- Changes in spinal cord

In gross deformity of the spine due to TB, unrelieved compression of the spinal cord results in loss of neurons and demyelination with progressive gliosis which might end in cord atrophy.

• Radiological Evaluation

In the paradiscal type of spinal TB, narrowing of the disc space is often the earliest radiological sign, this is commonly associated with loss of definition of the paradiscal margins of the vertebrae.The narrowing of the disc space is due to lack of nutrition with subsequent disc atrophy or due to prolapse of the nucleus pulposus into the soft necrotic vertebral body.

Reduction of the disc space is usually minimal or absent in the anterior and central types of spinal TB.The characteristic pathology is easily picked up by plain radiographs. In the anterior type, there is shallow erosion of the anterolateral vertebral bodies, while in the central type, concentric collapse might take place.

MRI IMAGING OF TUBERCULOUS SPINAL INFECTION

The intervertebral disc, rather than being destroyed, tends to herniate into the vertebral body. On MR imaging, the T2- weighted images demonstrate bony involvement and the prevertebral abscess as areas of increased signal. The prevertebral abscess may be larger than the degree of bony involvement. The disc signal may be normal or increased except at CI-C2 level where no disc is present. Gd-enhanced MR images will show enhancement of the paraspinal tuberculous abscess and the areas of bone involvement.

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 29, 2009 — One potential use for MR imaging in MS is to use test results to predict the subsequent development of the disease. This is most beneficial for patients with an atypical presentation or a typical but monophasic illness. A single episode of neurologic dysfunction is insufficient to make a diagnosis of MS no matter how typical or characteristic the symptoms and signs are. A number of studies now have evaluated the use of MR imaging for predicting the eventual evolution to CDMS both with variable clinical presentations or monosymptomatic features..

The subgroup of patients with MR imaging lesions consistent with MS at the onset of disease progressed to CDMS in 44% of the cases, whereas those with normal scans or scans not consistent with MS developed CDMS in only 5% . In the short term, MR imaging is quite sensitive for predicting development of clinically definite MS and conversely the lack of typical MS changes at the onset of disease is only infrequently associated with progression to MS.

• Monophasic Spinal Cord Disease

The risk of progression to MS with a complete transverse myelitis is relatively low, ranging from 2% to 15% .MS is much more commonly associated with what is called an incomplete or partial myelitis. Clinically, the patients present with incomplete mixed motor and sensory abnormalities that may or may not demonstrate a specific segmental level. Such partial myelitis syndromes may occur as the initial manifestation or during the course of MS.In general 59% of patients with acute spinal cord syndrome and with abnormal brain MRI scans consistent with CDMS at the onset of disease will develop CDMS whereas only 9% of patient with initially normal MR imaging develop CDMS.

• Chronic Myelopathies

The question of the predictive value in chronic myelopathies is somewhat less clear. Studies have indicated that 60% to 82% of patients with chronic progressive myelopathy will have cranial MR imaging abnormalities consistent with MS but pathologic confirmation studies have not been done. It is also unclear whether the 20% to 40% of patients with normal MR imaging have a different disorder or simply have MS involving only the spinal cord because occasional pathologically confirmed cases of MS restricted to the spinal cord have normal cranial MRI scans.

• Optic Neuritis

Similar findings have been reported with optic neuritis (ON) as for the acute partial myelitis syndromes. In patients with isolated ON, the reported rate of development of CDMS ranges widely between 15% and 75%. The presence of CSF oligoclonal bands predicts an increased risk of developing MS in this group of patients. Various reports indicate that lesions consistent with MS on MR imaging cranial scan can be found in 23% to 72% of patients with depending on the criteria used and that in the patients who have such lesions, there is a 55% to 70% risk of developing CDMS or LSDMS in follow-up ranging from 2 to 3 years. Conversely, patients who have isolated ON and a normal cranial MR imaging study have a much lower risk of developing multiple sclerosis with a range of 6% to 33% 4 years after onset.Thus for an individual patient with isolated optic neuritis, the risk for development of MS is 5 to 10 times greater in the presence of cranial MR imaging lesions consistent with demyelinating disease. Unfortunately, the absence of such lesions is not a guarantee that MS will not subsequently develop.

Imaging of the optic nerve itself was complicated initially by signal from orbital fat but by using STIR (short time inversion recovery) sequences to suppress the fat signal, MR imaging detects lesions in 84% of symptomatic patients although this value is less than that of visual evoked responses in the same patient group (100% sensitivity). Acute plaques in the optic nerves demonstrate gadolinium enhancement that correlates with visual dysfunction.

• Primary versus Secondary Progressive Multiple Sclerosis

Patients with a steady deterioration (progression) of clinical function may have such clinical progression from the onset of disease (primary chronic progressive ) or after an initially relapsing-remitting course (secondary chronic progressive ). Primary CP occurs more frequently, but not exclusively, in patients with onset after the age of 40. MR imaging studies in progressive patients has revealed significant differences in the appearances of primary CP versus secondary CP despite equivalent disability .Patients with primary CP MS have smaller lesions and additionally, new lesions enhance with gadolinium much less frequently. Spinal MR imaging in these 2 groups does not reveal significant differences. These findings are important with respect to clinical studies, most of which have not been stratified according to type of progression adding a further confounding variable in treatment trials.

• Acute versus Chronic Plaque

Much of the sensitivity of MR imaging in MS diagnosis relates to the fact that lesions persist in the brain despite the absence of any clinical symptomatology or resolution of prior clinical symptoms. Initially, MS plaques may appear quite large with rather fuzzy or indistinct borders but with resolution of the acute phase the lesions tend to contract in size and develop better definition between normal and abnormal tissue. This presumably reflects resolution of tissue edema and inflammation as the acute attack diminishes. The residual lesions seen on MR scanning weeks, months, or years later more likely represent demyelination, gliosis, and possibly enlarged extracellular spaces .Some early reports had described a resolution of MS plaques but generally these were studies done with low field strength magnets using a slice thickness of 10 mm, as compared to the more typical 5 mm slice thickness currently used. The studies using stronger field strengths and thinner slice thickness indicate that few plaques, defined as T2-weighted lesions, will disappear completely and those that do are almost always smaller than 5 mm initially .It is not uncommon, however, for new gadolinium-enhancing regions to disappear without leaving a T2-type abnormality

One feature of acute plaques is the disruption of BBB that previously had been detected using contrast-enhanced CT scans. With MR imaging, one can use paramagnetic gadolinium-containing chelates (such as gadolinium diethylenetriamine penta-acetic acid: GD-DTPA) that cross the BBB and provide contrast within the tissue. Gadolinium results in hyperintense signals in comparison to noncontrast studies due to shortening of TI. Gadolinium scans generally use T1-weighted parameters as the optimal modality for obtaining contrast images. Several studies have now used gadolinium enhancement to detect acute MS plaques and to define the characteristics of such plaques. In general, the enhancement in acute plaques persists for between 2 to 4 weeks .The acceptance of gadolinium enhancement as indicative of new or reactivated plaques has led to the use of gadolinium enhancement as an outcome measure for clinical therapeutic studies in MS.

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 29, 2009 — Studies using serial MR imaging scans have demonstrated that the degree of MR imaging activity significantly exceeds the degree of clinically detectable activity by a factor of 2 to 10.The excess of MR activity over clinical activity has a number of possible explanations. One explanation is that many lesions involve areas of the CNS that are not generally associated with any symptoms or signs. Alternatively, the lesions, as detected by MR imaging, may not indicate areas of tissue damage with functional abnormality. Areas of edema or inflammation related to BBB break- down may not be associated with axonal or neuronal dysfunction, features which would cause clinical symptoms, yet could still appear as large areas of abnormal signal on MR imaging. Nonetheless, this excess of MR imaging activity has important implications for clinical studies.If one uses only clinical attacks or deterioration as evidence of disease activity, one may overestimate or underestimate the benefit of a particular therapy. Clinical investigators are now moving much more towards the use of MR imaging results as secondary outcome measures for clinical trials and even as primary outcome measures .

As indicated previously, gadolinium enhancement appears to be a sensitive detector of new or newly active lesions in MS and a simple count of gadolinium enhancing lesions at frequent scanning intervals (4 to 8 weeks) may provide evidence of significant differences between treated and placebo groups before detecting differences in the numbers of relapses or deterioration on clinical rating scales.

In addition to counting new lesions appearing on MR imaging or counting gadolinium-enhancing lesions on MR, one can attempt to quantitate the total plaque volume within the nervous system. This method has been frustrated in part by lack of technology sophisticated enough to deal with the problem in a reliable fashion. Plaques often have borders that merge indistinctly with normal tissue and thus to assign volume to plaques is in good part arbitrary.Interobserver errors are quite significant with such techniques although intraobserver errors in well-trained observers can be quite acceptable. Such serial volumetric studies have indicated that plaque volume increases by about 20% over 2 years in both relapsing and chronic progressive disease .

• Clinical-MR Imaging Correlation

As experience with MR imaging in MS increases, apparently the degree of clinical dysfunction is not mirrored by findings on MR imaging scans. Severe clinical dysfunction can occur with minimal MR imaging changes or conversely large numbers of plaques could be detected by MR imaging in patients who were minimally symptomatic.The finding of severe clinical disability in the context of mild MR imaging changes is hard to explain unless one suggests that the lesions causing severe disability are below the resolution level of MR imaging scans. This argument becomes less tenable as the sophistication of machines increases and the findings persist. A more likely explanation is that the majority of the lesions causing dysfunction are not being visualized with head MR imaging but rather are in the spinal cord. Large numbers of plaques can occur in the spinal cord. A third possibility is that the disease under study is in fact not MS but another degenerative process with clinical features mimicking MS.

MS is much more active than is appreciated clinically, a finding hat has great implications for research activities in MS and further supports the use of MR imaging in clinical trials. MR imaging can not be used as the sole outcome measure, however, because clinical worsening can occur at times in the absence of MR imaging activity. One study found that despite combined cranial and spinal MR imaging, including gadolinium infusion, scans were considered active in only 93% of patients deemed clinically active.One aspect of correlating disability with MR imaging of the spinal cord is the technical difficulty of such imaging. In order to properly image the spinal cord one must use a surface coil rather than the coil of the MR imaging machine itself. The field strength of the magnet and the characteristics of the coil thus can have a significant impact on the quality of images obtained. The small size of the spinal cord further complicates imaging. Finally, the spinal cord tracts run in a vertical fashion and yet the plaques of MS may exist predominantly in a horizontal plane and may be detected with difficulty if sagittal images alone are made of the cord. Many plaques however, can extend over several segments and be readily detectable. Further complicating plaque detection in the spinal cord is the very close proximity of lesions to the CSE Using standard T1-weighted images, a signal from the CSF can obliterate the distinction of the increased signal of plaques from CSF By preceding the standard spin-echo pulse sequence for imaging the cord by an inversion pulse, however, one can obtain an image described as a FLAIR image.

Spinal MR imaging scans are most useful in patients with myelopathic presentation and in those older than the age of 50 with no lesions in the cerebrum or lesions that do not strongly suggest MS. Many times patients with clearly defined segmental levels have normal spinal MR imaging studies. This may be at least in part due to technical limitations as described above but nonetheless highlight the less than 100% sensitivity of such testing. Technological advances including fast spin-echo sequences in conjunction with multiarray coils greatly reduce the time needed for imaging the spinal cord. Spinal MR imaging virtually eliminates the need for invasive myelography. MR imaging studies can demonstrate both extra- and intra- parenchymal lesions thus providing information unavailable with standard CT myelography. The ability to scan the length of the spinal cord also provides an advantage over myelography if one is uncertain of the specific segmental level involved.

Although many reports of the lack of MR imaging and clinical correlation exist, some studies do report a correlation of clinical activity with the number and volume of gadolinium enhancing lesions seen on MR imaging although the correlation is complex.

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 28, 2009 — Medulloblastomas originate from poorly differentiated germinative cells of the roof of the fourth ventricle that migrate superolaterally to the external granular layer of the cerebellar hemispheres. Medulloblastoma may arise anywhere along the path of migration. This concept is useful to explain the fact that medulloblastomas in childhood are usually in the midline along the roof of the fourth ventricle, whereas those in young adults are nearly always located more laterally in the cerebellar hemisphere. Medulloblastomas constitute 25% of all pediatric intracranial neoplasms and approximately 40% of those arising in the posterior fossa.The peak incidence is in the first decade. Boys are two to four times more commonly affected. Medulloblastomas arising from the vermis may extend to involve the cerebellar hemisphere, brain stem, or fourth ventricle. From the fourth ventricle, they may extend into the cisterna magna via the foramen of Magendie, into the cerebello- pontine angle via the foramen of Luschka, or into the third ventricle via the aqueduct. Rarely large tumors may extend supratentorially and also inferiorly into the upper cervical canal. Owing to its close relationship with the fourth ventricle, medulloblastoma is frequently associated with hydrocephalus.

Figure 1. Gross and histopathological picture of medulloblastoma

On noncontrast CT, medulloblastomas generally appear as uniform high density lesions, which correlates well with histopathologic findings of a high degree of cellularity in this neoplasm. They tend to enhance homogeneously with sharp margins. Although approximately 53% of medulloblastomas have characteristic CT findings, a significant number of them have atypical features, including small cystic or necrotic areas, calcification, hemorrhage, lack of contrast enhancement, eccentric location, and direct supratentorial extension.

Figure 3.  Precontrast [left] and postcontrast [right] CT scan studies showing a calcified medulloblastoma with dense enhancement,notice the associated hydrocephalus

The MR appearance of medulloblastoma is variable . The tumor is typically hypointense to isointense to brain parenchyma on TI-weighted and isointense to hyperintense on T2-weighted images. The pattern of enhancement after intravenous injection of gadolinium is similar to that after injection of iodinated contrast material on CT. The greater sensitivity of MR imaging, however, often enables appreciation of a slightly heterogeneous enhancing pattern not as readily evident with CT. Subarachnoid metastatic seeding from primary medulloblastoma is well known.

Figure 4.  Pre,postcontrast MRI T1 (A,B,C) study showing a densely enhanced medulloblastoma,notice the associated hydrocephalus. D, MRI T2 image of the same patient

30% of patients with newly detected medulloblastomas have evidence of subarachnoid spread on the initial cranial CT study. Because gadolinium-enhanced MR is more sensitive than contrast-enhanced CT in detecting subarachnoid seeding, it is expected that this figure may be even higher.

Spread of medulloblastoma into the intracranial and spinal subarachnoid spaces and the ventricular system occurs more commonly than other pediatric posterior fossa neoplasms. If ventricles are shunted, seeding of tumor may occur at the other end of the shunt tube. For evaluation of recurrent or residual tumor, T2-weighted MR images should be obtained in conjunction with gadolinium- enhanced MR images because not all residual or recurrent tumors show contrast enhancement.

Figure 5. A  case of recurrent medulloblastoma, notice the meningeal and choroidal enhancement [A,B] and the spinal cord enlargement with focal enhancement [C,D]

Conversely, the presence of gadolinium-enhancement does not necessarily indicate the presence of residual neoplasm because radiation necrosis may present as areas of gadolinium enhancement.

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 28, 2009 —  Thyroid ophthalmopathy is characterized by inflammation, congestion, hypertrophy, and fibrosis of the orbital fat and muscles leading to enlargement of tissue, especially the extraocular muscles. The extraocular muscles are enlarged, firm, rubbery, and dark red. The histologic findings,that are related to the severity and stage of the disease, consist of interstitial edema and inflammatory cell infiltrates. The inflammatory cells are composed of lymphocytes, plasma cells, and occasional mast cells (B lymphocytes in 80% and T lymphocytes in 20%). The inflammatory reaction, predominantly in the muscle, can also occur in the tendons. The inflammatory reactions are predominantly localized in the endomysial connective tissue with extension to the perimysium and epimysium surrounding the extraocular muscles in the late stage. There is a fibroblast reaction with production of mucopolysaccharides, specifically, hyaluronic acid, characterized by glucose amino glycons.In the later stages of severe ophthalmopathy, there is fibrosis and fatty infiltrations of muscles resulting in a restrictive myopathy

• Extraocular muscles expansion

The diagnosis of Graves’ disease is often established by clinical means. CT is the preferred imaging modality for the evaluation of Graves’ disease in cases where the diagnosis is uncertain. Furthermore, CT is used to assess extra- ocular muscle enlargement, fatty expansion, and optic nerve compression, especially prior to surgery and as a follow-up after treatment. This examination should be carried out with 4- to 5-mm-thick sections in the axial and coronal planes. This allows detailed assessment of extraocular muscle enlargement, which provides the most important parameter in the diagnosis. The enlarged muscles are spindle- shaped with the belly reflecting the muscular portion and the tapered, anterior end, the tendinous portion. Occasionally, the muscular tendons may be slightly enlarged. On MR imaging, the normal muscle is characterized by low signal intensity on the Tl-weighted images and intermediate signal intensity on the T2- weighted images. There is marked enhancement of the extraocular muscles following the introduction of gadolinium, which is in contrast to muscles in other body parts, which reveal no enhancement. This is based on the increased vascularity of the extraocular muscles, allowing for diffusion of contrast material into the muscular tissue.

Figure 1. Two cases of gravis ophthalmopathy showing enlargement of the medial and lateral recti muscles,also notice exophthalmus

It has been shown that increased extraocular muscle volume correlates with severity of optic neuropathy and, furthermore, improvement of the optic neuropathy appears to correlate with decrease in extraocular muscle swelling at the apex of the muscle cone. Significant enlargement of the medial rectus muscle may lead to remodeling of the lamina papyracea with deviation medially from pressure by the medial rectus muscle.

Figure 2. A case of gravis ophthalmopathy showing enlargement of the extraocular muscles,notice that enlargement of the inferior rectus is only appreciated in the coronal cut, also notice exophthalmus more on the right side.

In Graves’ disease, a single muscle may be enlarged such as the medial , inferior and superior rectus muscles. If only axial images are performed, superior and inferior muscle enlargement is easily overlooked and is, therefore, optimally evaluated with coronal sections.Multiple muscles are usually enlarged in Graves’ disease and, not infrequently, both orbits are involved. Sometimes, the patient demonstrates clinical Graves’ disease in one orbit but on the CT study the asymptomatic orbit also reveals enlargement of muscles.

Figure 3. Showing two cases of gravis ophthalmopathy ,notice enlargement of the medial and lateral recti, also notice exophthalmus more on the right side

The margins of the muscles are usually well defined. In order of frequency, the inferior rectus is the most common muscle involved in Graves’ ophthalmopathy followed by the medial rectus muscle, and the superior muscle complex composed of the levator-palpebral, and superior rectus muscles. In addition, there is enlargement of the superior oblique muscle, which is optimally demonstrated in the coronal projection.

The lateral rectus muscle often reveals some enlargement in conjunction with the other extraocular muscles, but this is less pronounced and in most cases the muscle appears normal, whereas the medial and inferior rectus muscles reveal enlargement. The degree of muscle enlargement varies from mild to severe with a significant portion of the orbit obliterated when the muscles are markedly enlarged . Significant muscle enlargement leads to bunching in the apex of the orbit with extrinsic pressure on the optic nerve and consequent loss of vision and field defects.

• Expansion of the orbital fat

The second most common finding in Graves’ disease is expansion of the orbital fat. This is difficult to quantify on CT, but is suspected in patients with moderate to marked exophthalmos. Fatty expansion leads to considerable stretching and straightening of the optic nerve. Normally, the nerve is undulated when the globe is in a normal position. Frequently, there is a bulge of the orbital septum anteriorly secondary to extrinsic pressure from the expanded orbital fat. Occasionally, there are increased mottled densities within the orbital fat that, on biopsy, have proved to be lymphocytic infiltrations. Some vascular congestion may also contribute to a slight increase in soft tissue densities within the orbital fat, especially if they are linear in configuration. Slight enlargement of the lacrimal glands is also encountered in patients with Graves’ disease, which is well demonstrated on axial and coronal CT images. Rarely, there may be some slight enlargement of the optic nerve, probably the result of lymphocytic infiltrations in the surrounding orbital fat.

Several other disease entities may be responsible for enlargement of the extraocular muscles and, therefore, enter into the differential diagnosis; these include:

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 26, 2009 — The disturbance in the calcium and phosphorus metabolism that results from deficiency in the activity of the parathyroid gland produces a train of neurological problems that includes epilepsy, tetany, mental confusion, dementia, chorea, athetosis, torticolis, papilledema and others. Biochemically patients with hypoparathyroidism have reduced serum calcium level and elevated serum phosphorus level with inability to mobilize calcium from tissues in general and bone in particular.

CT scan of the brain in cases of hypoparathyroidism commonly reveals evidence of intracranial calcification that usually begins in the head of the caudate nucleus then extends to involve the rest of the basal ganglia.The pulviner and the cerebellar dentate nucleus are occasionally involved. The choroid plexus is also very frequently calcified.In advanced cases calcification might not be restricted to the subcorical grey matter and might involve the periventricular white matter Calcification is hyperdense on noncontract CT scan and appears as signal void structures on both the T1, T2 MRI images.

Figure 1. Precontrast CT scan study of a case of hypoparathyroidism showing calcification of the head of the caudate nucleus,putamen, choroidal plexus,thalamus [pulvinar] and the cerebellar dentate nucleus

Pathologically calcification commonly develop in and about the brain microvascular bed that supply the basal ganglia and other subcortical grey matter, presumably as a result of of some abnormality of the interstitial fluid that nourish the microvascular bed.  Although the serum calcium is low, however the existence of an excess of ionic calcium in the interstitial fluid explain the development of brain calcification in hypoparathyroidism.

Figure 2. Bone windows of figure [1]

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 26, 2009 — Wilson disease is an uncommon autosomal recessive disorder of copper metabolism characterized by pathological changes mostly affecting the liver and brain.In the brain pathological changes mostly affect the basal ganglia,thalamus and cerebellum.

MRI has proved to be an excellent tool in studying cases of Wilson disease. In the heavily weighted T2 images the following changes are observed.

These changes are commonly bilateral and symmetrical

Such unusual signal may be explained by iron deposition in areas which has excessive copper accumulation. It has been observed that siderophages are commonly observed in relation to cavitations observed in the basal ganglia in the late stages of Wilson disease.

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 25, 2009 —  Ependymomas are gliomas arising from the ependymal cells, usually within the ventricles of the brain or central canal of the spinal cord. Less frequently they may occur in the brain parenchyma. Intracranial ependymomas are usually found in children, whereas intraspinal ependymomas are more frequently seen in adults. The majority of intracranial ependymomas are infratentorial in location (70%), particularly in the fourth ventricle.

Ependymomas occur more commonly during the first 5 years of life. Both sexes are equally involved. Two histopathological types exist, the cellular type and the papillary type

THE CELLULAR TYPE

THE TUMOURS ARE HIGHLY CELLULAR AND COMPOSED OF POLYGONAL CELLS AND LITTLE SUPPORTING STROMA. TWO ARCHITECTURAL FEATURES ARE FOUND IN THE CELLULAR TYPE

· EPENDYMAL TUBULES [EPENDYMAL ROSETTES] :- COMPOSED OF CONCENTRIC ARRANGEMENT OF CILIATEED EPENDYMAL CELLS AROUND A GENUINE CAVITY

· PERIVASCULAR PSUEDOROSETTES :- PERIVASCULAR ARRANGEMENT OF EPENDYMAL CELLS FORMING PSUEDOSETTES

Figure 1. The cellular type

THE PAPILLARY TYPE

TWO TYPES ARE FOUND

· THE PAPILLARY EPENDYMOMAS:-THE EPENDYMAL CELLS RESTS UPON GLIAL FIBRILLAY STROMAS

· MYXOPAPILLARY EPENDYMOMAS:-THE CONNECTIVE TISSUE STROMA IS THE SEAT MYXOMATOUS DEGENERATION

On CT scans, ependymomas show variable density, including hypodense, isodense, hyperdense, and mixed density patterns. Most ependymomas show patchy enhancement. Occasionally they enhance uniformly, and a small percentage exhibit no enhancement. A rim of cerebrospinal fluid within the fourth ventricle surrounding ependymoma can be appreciated in roughly one half of CT studies. Ependymomas may extend through the foramen Luschka and Magendie into the vallecula, foramen magnum, cerebellopontine angle, and upper cervical subarachnoid space. A tongue of tumor mass extending into the foramen magnum is characteristic for ependymoma, although this finding may also be seen with medulloblastoma. Ependymoma, similar to medulloblastoma, is often located within the fourth ventricle and consequently is frequently associated with hydrocephalus.

Calcifications are seen in approximately 50% of pediatric posterior fossa ependymomas. Calcification in ependymomas ranges from a small nodular focus to large nodules. Although calcification in a childhood posterior fossa neoplasm might suggest the diagnosis of ependymoma, one must keep in mind that other statistically more common tumors may also calcify. Calcification is seen in only 13% of medulloblastomas, however, and in only a small percentage of cerebellar astrocytomas.

MR shows the solid component of the tumor to be hypointense to isointense to brain parenchyma on TI-weighted images and hyperintense on T2-weighted images. It is usually heterogeneous in signal intensity owing to the presence of calcification, small cystic areas, hemorrhage, and vascularity or a combination of these characteristics. The cystic component of ependymoma tends to show slightly increased signal intensity as compared with cerebrospinal fluid owing to its increased protein content. Heterogeneous enhancement is seen after the intravenous injection of gadolinium.

	Figure 2. MRI T1 postcontrast study [A,B] and T2 image [right image] showing a densely enhanced cerebellar ependymoma,notice the signal heterogeneity which could be due to calcification or blood products.Also notice that the 4TH ventricle is pushed posteriorly and compressed

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 25, 2009 —  Although they may be located in diverse anatomic regions, the glomus organs of the human body are homologous organ systems with anatomic and functional similarity. Paraganglioma, glomerocytoma, and chemodectoma are terms used synonymously for tumors of these organ systems.

Characteristic of this group of tumors is their high grade of vascularization; the entire organ is interlaced with an arterial vessel labyrinth that drains into wide venous vessels, situated predominantly on the surface of the tumor. The different locations of tumors of the paraganglion tissue result in the extraordinary variability in presenting clinical findings. Nonspecific otologic symptoms, including tinnitus and conductive hearing loss, must be further evaluated with audiometry and a neuro-otologic examination.

• Pathology of glomus tumors

ANGIOGRAPHY OF GLOMUS TUMOURS

Prior to the advent of CT, selective angiography of the external carotid and vertebral arteries was used to establish the diagnosis, to assess the grade of vascularization, and to identify feeding vessels and associated vascular anomalies. In the early stages, glomus tumors receive their blood supply from branches of the external carotid artery. Advanced tumors gain supply from the vertebral and basilar arteries. Angiography used to be important for visualizing surgically important variants of vessels, such as the laterally displaced internal carotid artery in its petrous course, or a cranially localized jugular venous bulb.

• Angiography of glomus tumors: Often have the characteristics of arteriovenous malformations

CT SCAN IMAGING OF GLOMUS TUMOURS

CT is currently performed to categorize glomus tumors and to depict foramenal expansion. High- resolution CT can reveal the intratympanic and intracranial expansion of the tumor and clarify its relationship to the cervical soft tissues and to identify an aberrant carotid artery, or high jugular bulb. In most cases, glomus tumors can be differentiated from other skull base processes rising dynamic CT with a time- density profile.

However, CT produces a relatively high ratio of false-negative and false-positive results, especially when performed only after intravenous administration of contrast agent. Small tumors, especially those originating from the tympanic glomus, render diagnosis more difficult because they present only as hypointense areas without osseous destruction. MR imaging, dynamic MR, and MRA, with their superior soft-tissue contrast, have since become the diagnostic modalities of choice for evaluating these tumors.

Figure 3.  Plain x ray [left] and CT scan bone window [middle ,right] showing extensive destruction of the jugular foramen in a case of glomus tumour

• CT scan characteristics of glomus tumors

• Classification of Glomus Tumors.

For pretherapeutic planning, glomus tumors must be differentiated according to their origin and location. Although several classification systems exist for staging these tumors, the Valavanis and Fisch classification system is most widely established. [1]

Type A tumors represent the glomus tympanicum tumors at the cochlear promontory, whereas hypotympanic tumors are classified as type B tumors, which erode the hypotympanic osseous structures. Characteristically, the cortical border of the jugular bulb is not affected. Type C tumors are defined as jugular glomus tumors without intracranial expansion and are subclassified in types C1 to C4 according to the extent of osseous involvement. Type D glomus tumors are glomus jugulare tumors with intracranial expansion and either extradural or intradural tumor spread.

MRI IMAGING OF GLOMUS TUMOURS

Several factors make primary evaluation of glomus jugulare tumors with MR imaging advantageous. MR imaging is superior to CT because of its superior soft-tissue contrast in the absence of signal from surrounding bone tissue. Flow phenomena allow visualization of flowing blood, and are a further reason for the superiority of MR imaging. The topographic relationship of the carotid siphon and the jugular bulb can be reliably determined on unenhanced MR images. MR imaging also allows differentiation between glomus tumors and vascular anomalies, the most frequent of which are the aberrant internal carotid artery and the cranial jugular bulb.

The aberrant internal carotid artery is characterized by atresia from the carotid bifurcation to the petrous segment, with collateralization of the atretic segment over the ascending pharyngeal and the caroticotympanic arteries.The cranial jugular bulb is caused by a diverticular bulge of the superolateral part of the jugular bulb into the hypotympanic cavity. Such vessel variants can be clearly differentiated from glomus tumors extending into the jugular vein after administration of GD-DTPA, particularly in coronal slice orientation or when using MRA. Retrograde phlebography of the internal jugular vein has thus become largely obsolete.

In some patients with small glomus tumors, plain MR imaging reveals at best limited diagnostic information. In these cases, fast imaging techniques with GD-DTPA facilitate the diagnosis. Similar to the time-density profile obtained in dynamic CT, the signal intensity of glomus tumors can be plotted over time when analyzing fast gradient-echo sequences. In all untreated glomus tumors, a rapid and high increase in signal intensity can be observed during the first 60 seconds after administration of GD-DTPA, with an average enhancement factor of 2.5. Enhancement reaches its maximum after 120 to 160 seconds, decreasing approximately 300 to 350 seconds after administration of contrast.

This pattern of signal intensity after injection is due to the high degree of vascularization of these tumors and correlates with the findings of dynamic CT. In comparing the signal intensities of plain and enhanced sequences, an average GD-DTPA uptake of up to 205% is seen in tumors; enhancement is significantly lower in muscle (23%) and fatty tissue (51%).

Analysis of mean enhancement and enhancement-time curves facilitates the diagnosis of glomus tumors and contributes to the differential diagnosis. Characteristically, the analysis of the dynamic series in glomus tumors demonstrates the dropout effect in form of a dip in the time-intensity curve with high dose contrast material injection. This characteristic dropout effect can be observed in all glomus tumors, independent of their location. size, or classification .In dynamic CT, scans with time-density profiles do not show this characteristic dropout effect and lesion differentiation renders difficult.

In most patients, optimal imaging with clear demarcation of the vascularized parts of the tumor is achieved by administration of GD-DTPA, although in some patients with large tumors no additional information is obtained with contrast enhancement. In several comparative analyses, MR imaging was found to be superior to CT in diagnostic accuracy, determination of topographic orientation, and demarcation of glomus tumors. Especially for processes infiltrating the middle cranial fossa (coherent mastoiditis, small glomus tumors), enhanced MR imaging shows its superior diagnostic potential. All patients with glomus tumors of the skull base or the temporal bone must undergo an angiographic study (digital subtraction angiography [DSA] or MRA). Evaluation of the vascular supply reveals important additional information regarding the vascularization of the tumor and the hemodynamic situation in the Circle of Willis. Other glomus tumors are seen to be supplied by the ascending pharyngeal artery.

The high overall reliability of MR imaging is based on its capacity to reveal other tumors of the temporal bone and the cerebellopontine angle. Schwannomas and meningiomas can be readily identified by their typical topographic sites and the characteristic temporal characteristics of signal intensity after administration of GD-DTPA. In contrast to glomus tumors, these tumors show a slower increase in signal intensity during the early phase and a constant increase up to the seventh minute after injection.

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]

The author: Professor Yasser Metwally

http://yassermetwally.com

INTRODUCTION

June 19, 2009 —  The pituitary gland lies within the sella turcica between the cavernous sinuses . Its density is similar to that of the sinuses and dura so that, with the possible exception of its upper surface, which is to a variable degree outlined by the chiasmatic cistern but partly covered by the pituitary diaphragm, the precise limits of the gland cannot be distinguished from the adjacent tissues on either plain or contrast-enhanced studies.

The shape and height of the pituitary gland is best assessed on the coronal views. The height should be less than 8 mm. The top of the gland should be flat or concave, and there should not be an upward convexity contour.The normal pituitary appears slightly hyperdense on the plain scan, and there is homogeneous contrast enhancement.

• Pituitary microadenoma

These tumors may be 3 to 10 mm in size and may be located within a normal-sized sella turcica. They may cause symptoms of hormonal hypersecretion. These are most commonly caused by prolactin or growth hormone abnormalities, less commonly by adrenocorticotrophic hormone disturbances.The elevated pituitary hormone content may be caused by conditions other than pituitary neoplasms; therefore, sensitive neuroimaging studies are necessary to document the presence or absence of pituitary microadenomas

Because of the small size of pituitary microadenomas, the measured sella volume may be within normal limits; however, even with normal size of the sella, the sellar shape and bone detail almost always show some detectable radiographic abnormalities. This may not always be detected by routine skull radiographs (or even utilizing coned-down views of the sella turcica), and these abnormalities may most sensitively be assessed by CT scan with a bone windows. The most characteristic radiographic abnormal finding of pituitary microadenomas is an anterior-inferior bulge in the sella floor.

This is most commonly seen in the lateral wall of the sella, correlating with the previously reported propensity of prolactin-secreting microadenomas located in the lateral portion of the pituitary gland. It has been reported that computerized tomography shows sella turcica bone abnormalities in 96 per cent of pituitary microadenomas. However, it is also important for the clinician to understand the pattern of normal variations in the development of sella turcica and the contiguous sphenoid bone. This understanding may avoid interpretative errors in assessing pituitary radiographic changes as being caused by tumor when these changes may actually be due to normal anatomic variants.

The CT findings that are suggestive of a pituitary microadenoma include (1) height that exceeds 8 mm with an upward bulging or a convexity to the superior surface of the gland, (2) focal hypodense lesion seen within the hyperdense gland, (3) upward and lateral deviation displacement and enlargement of the pituitary stalk or infundibulum . If the infundibulum (as seen on the axial section) is larger than the basilar artery (located in the interpeduncular cistern) on the enhanced scan. this is considered to be abnormal, and this finding is suggestive of a pituitary mass. The upward extension and displacement of the infundibulum due to a pituitary tumor is best seen on the coronal views.

Figure 2.  Postcontrast CT scan showing pituitary microadenoma demonstrated as rounded relatively hypodense masses in the lateral portion of the pituitary gland ,notice loss of the normal upward concavity [now there is upward convexity] of the gland and destruction of the sellar floor [arrows]

The prolactin-secreting microadenomas are equally distributed between central and lateral location within the gland; whereas growth hormone and adrenocorticotrophin-secreting microadenomas are usually more central in location. After infusion of contrast material, the microadenoma enhances more slowly than the normal pituitary gland. This results in the focal hypodense appearance of the microadenoma . if the postcontrast scan is delayed, the focal hypodensity representing the microadenoma may not be seen. Following treatment with bromocriptine, the shrinkage in the size of the pituitary mass may be well followed with serial CT.

Utilizing high-resolution computed tomography, it is possible to detect pituitary microadenomas in most cases. A complete CT scan study must include direct coronal sections that are 1.5 to 2.0 mm in thickness. However, reformatted reconstructions (which are based upon the axial views and are then generated into the coronal and sagittal planes by computer analysis) may be utilized.

MRI is more sensitive than CT scan in detecting pituitary microadenomas.It is best demonstrated on the postcontrast T1 images as a rounded hpointensity that shows significant delay in enhancement compared with the normal pituitary gland tissues.

• Pituitary macroadenomas

The CT findings in pituitary macroadenomas are dependent upon several factors. These include size of tumor, major vector of expansion, and tumor pathologic characteristics. If the pituitary adenoma is a solid tumor , it usually appears iso- or hyperdense (noncalcified) on the noncontrast CT, and there may be dense homogeneous sharply marginated contrast enhancement. Cystic adenomas appear as round hypodense lesions on the noncontrast CT scan, and there is usually a thin peripheral rim of enhancement. In rare instances, the cystic pituitary adenoma appears as a hypodense lesion without contrast enhancement. Hemorrhagic pituitary adenomas usually appear as hyperdense noncalcified lesions on the plain scan; there is dense homogeneous or peripheral rim enhancement.

Figure 4.  Postcontrast CT scan showing pituitary macroadenoma with suprasellar and infrasellar extension into the sphenoidal sinus

If the pituitary neoplasm ,as demonstrated by CT scan contains necrotic liquefied tissue rather than solid hematoma, the plain scan may show a more mottled hypodense central region with a peripheral rim of enhancement.Invasive adenomas may appear as irregularly marginated hyperdense lesions; they may show heterogeneous enhancement . They are diffuse, widespread, and poorly marginated lesions; they also show marked bone erosion.The presence of intrasellar calcification should suggest an alternative diagnosis such as craniopharyngiomas, meningiomas, aneurysms; however, in rare instances, pituitary adenomas show evidence of calcification.

Figure 5. Plain x ray [left] and CT scan bone window [coronal view] showing double flooring of the sella due to unilateral depression of the sellar floor [left] induced by a pituitary adenoma

Because pituitary adenomas usually originate within the sella turcica, CT shows an enhancing round mass. There is usually no surrounding suprasellar cistern may be seen on axial sections; however, these tumors are more clearly defined on coronal and sagittal sections. The superior (extending to the intraventricular foramina and anterior third ventricle) and inferior (into the sphenoid sinus) extension of the mass is best demonstrated with coronal CT.

Figure 6.  Postcontrast CT scan [left] and plain x ray sella [right] showing pituitary macroadenoma with suprasellar extension inducing ballooning of the sella [right]

The sphenoid sinus is located directly underneath the floor of the sella. Tumor extension into the air-filled sinus and evidence of bone erosion of the sella floor is well visualized on coronal CT. Lateral extension of the pituitary adenoma may be demonstrated by displacement of the carotid arteries, which are paired structures located in the antero-lateral portion of the suprasellar cistern.

Figure 7.  Postcontrast CT scan [left] and plain x ray sella [right] showing pituitary macroadenoma with suprasellar extension inducing ballooning of the sella [right]

Figure 8.  Plain x ray sella showing extensive destruction of the sellar floor by a highly invasive pituitary adenoma

The cavernous sinuses in the parasellar region appear as paired symmetrical vertically oriented densely enhancing parasellar bands. With lateral extensions of the adenoma,the cavernous sinus appears as a broad band that is thicker ipsilateral to the tumor. The asymmetry or lateral deviation of the broad band of cavernous sinus enhancement is consistent with lateral extension of the intrasellar mass. Anterior extension of adenomas is demonstrated by the presence of an enhancing mass located within the anterior portion of the suprasellar cistern. With more significant anterior extension, there are enhancing lesions in the frontal region seen with surrounding hypodensities. If there is posterior extension, there is distortion and posterior displacement of the interpeduncular cistern and basilar artery. Rarely, pituitary adenomas extend to the intraventricular foramina to cause obstructive hydrocephalus; however, this finding is more common with suprasellar masses such as craniopharyngiomas.

Table 1. Grading of pituitary adenoma

Figure 9.  MRI T1 study showing a huge pituitary adenoma with suprasellar extension

Enlargement of pituitary adenomas during pregnancy is well documented and may be demonstrated by CT. Rarely hypopituitarism can develop in previously normal women during pregnancy or the postpartum period associated with extensive infiltration of the gland by lymphocytes and plasma cells, referred to as lymphocytic hyophysitis. CT reveals sellar enlargement by a homogeneously enhancing mass bulging into the suprasellar region .

Originally termed chromophobe adenomas, endocrine-inactive pituitary tumors were once considered the largest group of pituitary tumors . With advances in endocrinologic testing and modern immunohistochemical and immunoelectronic microscope techniques, the incidence of adenomas with no evidence of hypersecretion or endocrine activity has decreased to about 25 per cent of pituitary adenomas. Histologically, these adenomas have secretory granules and immunocytochemically are growth hormone or prolactin-positive, despite no associated clinical changes or abnormal serum hormone levels about 5 per cent of the time. Inactive tumors have cells with no histologic, immunocytologic, or electron microscopic markers (Null cells). They are chromophobic and electron microscopy show them to have poorly developed cytoplasm, indented nuclei, and sparse granules (100 to 250 lim) lined up along the cell membrane.

It is the functionally active group of pituitary tumors that comprise the largest percentage of pituitary adenomas. They represent about 75 per cent of all pituitary tumors. Preoperative endocrinologic testing, as well as clinical symptomatology resulting from the adenoma’s hypersecretion of hormones, helps to identify and classify these tumors. It is this functional classification confirmed with immunohistochemical and immunoelectromicroscopic techniques and not traditional light microscopic pathology that separates these tumors.

Prolactinomas represent about 40 to 50 per cent of all patients with pituitary adenomas.Under light microscopy, prolactin cell tumors are chromophobic or acidophilic. Using immunoelectron microscopy, they may be classified as densely or sparsely granular, although the former type is quite rare. The densely granular resemble nontumor lactotrophic pituitary cells that are resting and nonsecreting. The sparsely granular type resemble the nontumor lactotrophic pituitary cells that are actively secreting. Their secretary granules are sparse, spherical, and measure 150 to 350 nm.

Somototrophic adenomas, resulting in acromegaly, account for 15 to 25 per cent of pituitary adenomas. Under light microscopy, these tumors may be termed acidophilic or chromophobic. Using immunoelectron microscopy, two distinct cell types can be identified: densely and sparsely granulated adenomas. The densely granulated cell type more closely resembles nontumor pituitary somatotropic cells and is characterized by well-developed endoplasmic reticulum, permanent Golgi complexes, and numerous spherical densely staining secretary granules. The sparsely granulated type differ from nontumorous pituitary somatotropic cells in that it has permanent Golgi complexes, irregular nuclei, few spherical secretary granules, and several centrioles.

Cushing’s disease or Nelson’s syndrome caused by corticotropin-secreting adenomas represent only about 5 per cent of all pituitary adenomas. Under light microscopy, corticotrophic cells are basophilic. Immunoelectron microscopy shows these tumor cells to be similar to corticotrophic nontumorous pituitary cell types containing numerous spherical secreting granules that vary in density, measure 250 to 700 nm, and line up along the cell membranes.

The rarest of pituitary adenomas are those that secrete solely thryotrophin or gonadotropin. Each type accounts for less than 1 per cent of pituitary adenomas. Under light microscopy, the thyrotropic adenomas are chromophobic and under electron microscopy, they have long cytoplasmic processes, sparse, spherical secreting granules (150 to 250 nm), and abundant endoplasmic reticulum.

• Pituitary apoplexy

Pituitary apoplexy is due to infarction of or haemorrhage into a pituitary adenoma. Infarction may be indistinguishable from a low density pituitary swelling and may or may not show enhancement. Haemorrhagic pituitary apoplexy may reveal high density within the adenoma or brain substance or subarachnoid space in the acute phase and low density with or without marginal enhancement as the hematoma is absorbed. This condition will probably be considered by the clinician when an appropriate syndrome occurs in a patient known to have a pituitary adenoma, but pituitary tumours may first present as subarachnoid haemorrhage. The correct diagnosis should be recognized from CT or suspected from sellar erosion on plain films prior to neuroimaging studies [CT scan or MRI].

Pituitary apoplexy commonly results in spontaneous involution of the pituitary adenoma and if the patient survives, this might result in empty sella

• Empty sella syndrome

In patients with radiographic and polytomographic evidence of an abnormal sella turcica, it is important to differentiate a pituitary mass lesion, such as pituitary macroadenomas, intrasellar cysts, intrasellar aneurysms, from intrasellar cisternal herniation (an empty sella).In the empty sella syndrome, the sella turcica is enlarged, usually with none or only minimal bone erosion; however, bone erosion identical to that seen in pituitary neoplasms-may be seen in some cases .In the empty sella, the pituitary gland is flattened and atrophic; it is located in the posterior-inferior portion of the sella turcica. CT shows evidence of CSF-density extending into the sella turcica on both the coronal and sagittal views.

Figure 11.  CT scan showing intrasellar extension of the suprasellar cistern in a case of empty sella inducing ballooning of the sella [left]

There is no evidence of abnormal intrasellar enhancement. With thin section CT, the pituitary infundibulum may be seen extending downward into the sella. This is the most important point in differentiating an empty sella from a pituitary adenoma.In some cases, the diagnosis of an empty sella may only be established with metrizamide CT cisternography.The diagnosis is established by the finding of opacification of the intrasellar cistern.Metrizamide CT cisternogram is frequently necessary to differentiate an intrasellar subarachnoid cyst or a pituitary micro- or macroadenoma from an empty sella . It is important to be aware that surgically proved hormonally secreting pituitary microadenomas have occurred in patients with CT evidence of an empty sella

Figure 12.  CT scan [left] and MRI T1 [right] showing intrasellar extension of the suprasellar cistern in a case of empty sella

This may complicate a pituitary tumour or occur in the presence of a microscopically normal pituitary gland. The first type may follow surgery or therapy for pituitary neoplasm.

In patients with a deficient pituitary diaphragm, intrasellar extension of the chiasmatic cistern may cause enlargement of the sella turcica and compress the normal pituitary gland to the periphery of the enlarged sella. Such patients are usually discovered when a skull radiograph is taken for investigation of an unrelated condition such as non-specific headache or trauma. The sella is usually symmetrically enlarged and commonly disproportionately deep or quadrangular in shape, although it may be asymmetrical or ballooned and thus simulate a pituitary tumour. High resolution thin CT sections of the pituitary fossa will show that the sellar contents are of CSF attenuation; the infundibulum can usually be traced lying closer to the dorsum than the anterior wall of the sella and extending down to the thinned pituitary gland, sometimes as little as I mm in depth, lying adjacent to the floor. The appearances are confirmed by coronal and sagittal reformatting. If head scanning shows no additional abnormality further investigation is contraindicated.

However, in a patient with deficiency of the Pituitary diaphragm empty sella may be a complication of raised intracranial pressure It is most commonly associated with pseudotumour cerebri and therefore in obese or hypertensive women, but sometimes with convexity block to CSF flow and with intracranial tumours. In such conditions visual field defects and visual loss may be caused by intrasellar herniation of the optic chiasm or nerves, and erosion of the walls of the sella may result in a fistula into the sphenoid air sinus, causing CSF rhinorrhoea and/or fluid in the sinus.

Pituitary apoplexy is due to infarction of or haemorrhage into a pituitary adenoma. Infarction may be indistinguishable from a low density pituitary swelling and may or may not show enhancement. Haemorrhagic pituitary apoplexy may reveal high density within the adenoma or brain substance or subarachnoid space in the acute phase and low density with or without marginal enhancement as the hematoma is absorbed.

This condition will probably be considered by the clinician when an appropriate syndrome occurs in a patient known to have a pituitary adenoma, but pituitary tumours may first present as subarachnoid hemorrhage.The correct diagnosis should be recognized from CT or suspected from sellar erosion on plain films prior to angiography. Pituitary apoplexy is one cause of spontaneous regression of pituitary adenoma and of empty sella.

References

1. Metwally, MYM: Textbook of neuroimaging, A CD-ROM publication, (Metwally, MYM editor) WEB-CD agency for electronic publication, version 10.2a April 2009 [Click to have a look at the home page]