Show Summary Details
Page of

Imaging of the pituitary 

Imaging of the pituitary
Chapter:
Imaging of the pituitary
Author(s):

J. V. Byrne

DOI:
10.1093/med/9780199235292.003.2070
Page of

PRINTED FROM OXFORD MEDICINE ONLINE (www.oxfordmedicine.com). © Oxford University Press, 2021. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 28 September 2021

Imaging methods

MRI is the optimum method of imaging the pituitary of patients with suspected pituitary disease though CT is an acceptable alternative. The advantages of MRI are: direct multiplanar scanning, lack of ionizing radiation, and good anatomical tissue discrimination. Imaging the pituitary gland and hypothalamus is best performed in the sagittal and coronal planes because they show the relationships between gland and adjacent structures. Scanning in the axial plane alone is a poor technique for demonstrating vertical relationships between structures lying between the floor of the third ventricle and sella turcica. Computer-generated three-dimensional (3D) reconstructions of axially acquired data (by MR or CT) and can be viewed in any plane can compensate but usually direct scanning gives better resolution images.

The disadvantage of MRI in this situation is its relative insensitivity to pathological calcification, and lack of signal from corticated bone. CT or even plain film radiography may be required to demonstrate or exclude pathological calcification. In this respect CT is far more sensitive than plain film radiography and the only remaining role for the latter in pituitary imaging, is to exclude metallic implants that might be contraindications to MRI. For surgical planning and intraoperative guidance some surgeons prefer the level of bony detail that 3D CT images provide (1). Conventional catheter angiography is rarely indicated because both MR and CT angiography (MRA, CTA) are capable of identifying the positions of the intracavernous and supraclinoid carotid arteries and differentiate pituitary mass lesions from aneurysms. Very rarely the diagnoses of a substantially thrombosed aneurysm requires intra-arterial digital subtraction angiography. Angiography continues to have a role during catheter navigation for venous sampling in patients being investigated for causes of Cushing’s syndrome.

MRI techniques

Various technical refinements to pituitary MRI have been advocated but given the enormous number of potentially useful MRI sequence protocols’, basic scanning methods are remarkably similar in different centres. It is generally agreed that the structures of the sella region are best imaged using T1-weighted sequences which are constructed to produce images with dark cerebrospinal fluid (CSF), grey brain and white fat. Corticated bone returns low signal and appears dark but bone marrow fat returns high signal and therefore appears white. The pituitary gland returns signal similar to cerebral white matter and flowing blood little or no signal and therefore is black. The latter is the basis for MRA (see Fig. 2.3.5.1).


Fig. 2.3.5.1 (a) Sagittal and (b) coronal T1-weighted MR images showing the normal pituitary gland and stalk (arrow). In the sagittal view the posterior lobe returns hyperintense signal at the site of antidiuretic hormone storage (short arrows).

Fig. 2.3.5.1
(a) Sagittal and (b) coronal T1-weighted MR images showing the normal pituitary gland and stalk (arrow). In the sagittal view the posterior lobe returns hyperintense signal at the site of antidiuretic hormone storage (short arrows).

The anatomical relationships of the infundibulum of the hypophysis and hypothalamus are easily identified. Using T1-weighted sequences the nuclei of the hypothalamus cannot usually be distinguished. Areas of bright signal in the stalk and posterior lobe are evident in up to 50% of T1-weighted scans performed in patients without endocrine diseases. The observed frequency declines with increasing age (2) and without this chemical difference in magnetic property the anterior and posterior lobes of the gland cannot be readily distinguished. The effect identifies the site of antidiuretic hormone storage (Fig. 2.3.5.2), and has been variously ascribed to vasopressin, neurophysin, or phospholipid vesicles in the neurohypophysis (3).


Fig. 2.3.5.2 Sagittal MR sequence showing a small aneurysm of the right internal carotid artery which points medially and extends into the sella (arrow). Flowing blood does not return signal and so appears black on MRI. Note the how the parent and contralateral carotid arteries also appear black on this sequence.

Fig. 2.3.5.2
Sagittal MR sequence showing a small aneurysm of the right internal carotid artery which points medially and extends into the sella (arrow). Flowing blood does not return signal and so appears black on MRI. Note the how the parent and contralateral carotid arteries also appear black on this sequence.

The power of MRI to resolve different structures depends on the signal-to-noise ratio (SNR) of the acquired signal. A high SNR is required to detect small abnormalities and small anatomical structures. Simply increasing the matrix size, i.e. the number of pixel (or voxel) elements will not improve the SNR though decreasing the pixel size will improve the spatial resolution of the image. To improve the SNR and spatial resolution it is necessary to increase the magnet gradient strengths and to acquire more data by lengthening the scan time (3). Pituitary MRI is therefore best performance in high field strength imagers using longer echo time (TE) sequences (4). In practice T1-weighted spin echo sequences are performed with repetition times (TR) of 500–600 ms, echo times (TE) of 15 ms and two or more excitations. Scanning is performed in coronal and sagittal planes using a minimum matrix size 256 × 256, to give 3 mm thick contiguous slices. Typically scanning takes 5–8 min at 1.5 T for each sequence. An alternative approach is to use a T1-weighted gradient echo technique with a 3D Fourier synthesis so that subsequent computer manipulation allows the imaged sample to be viewed in any plane and the reader can review and clarify any suspicious areas. Some centres perform complementary T2-weighted sequences in order to use the high signal returned from the CSF to outline structures in the chiasmatic and other basal cisterns and to assess signal return from tumours (Fig. 2.3.5.3).


Fig. 2.3.5.3 (a) Coronal T1-weighted and (b) axial T2-weighted MR images showing a macroadenoma invading the right cavernous sinus. The tumour is also extending into the chiasmatic cistern but not the left cavernous sinus. Note how the tumour surrounds the right carotid artery and returns a mixed signal on the T2-weighted sequence.

Fig. 2.3.5.3
(a) Coronal T1-weighted and (b) axial T2-weighted MR images showing a macroadenoma invading the right cavernous sinus. The tumour is also extending into the chiasmatic cistern but not the left cavernous sinus. Note how the tumour surrounds the right carotid artery and returns a mixed signal on the T2-weighted sequence.

The IV administration of paramagnetic agents such as gadolinium, which are taken up by the gland and surrounding tissues is useful but its routine use is controversial. These agents, like radiographic contrast media used in CT scanning, do not cross the blood–brain barrier. The pituitary gland and stalk therefore enhance and appear whiter on T1-weighted images. The hypothalamus and optic chiasm do not enhance if the blood–brain barrier is intact. Blood vessels, meninges and mucosa of the paranasal sinuses will enhance. The role of gadolinium-enhanced MRI in the investigation of different pathologies will be considered below. Dynamic MRI has been used to study the timing of intravenously administered gadolinium uptake by the pituitary gland. Obtaining rapid single slice images, Sakamoto et al. (5) demonstrated that the stalk and posterior lobe enhanced 20 s after an intravenous injection and that this extended into the anterior portion of the gland within 80 s. Using this technique can increase the detection rate of microadenomas and is used in patients with suspected Cushing’s disease and apparently normal glands on conventional scanning (6).

CT scanning techniques

Apart from its complementary role in detecting pathological calcification, CT scanning remains the primary imaging modality for the small proportion of patients who are unable to undergo MRI. Patients who are extremely claustrophobic, have cardiac pacemakers or other implants such as intracranial aneurysms clips and traumatic metallic fragments, which are sensitive to the effects of the magnetic field, cannot be scanned. CT is then used and multislice imagers with helical scanning are able to acquire axial images in less than a minute. These are displayed in 2D or 3D views by computer post processing. Currently available scanners generally produce images of sufficient quality to demonstrate sella anatomy on unenhanced images but intravenous injection of iodinated contrast media is generally used to improve tissue contrast. It is taken up by the hypophysis in the same way as the MRI contrast agent, gadolinium. Thus microadenomas, craniopharyngiomas, and tumours of the hypothalamus enhance and are better delineated but demonstration of microadenomas within a morphologically normal pituitary depends on differential uptake rates.

CT, unlike MRI, cannot demonstrate blood vessels without the injection of contrast agents. CTA is performed during ‘first-pass’ after an intravenous bolus injection of radiographic contrast media. It requires accurate timing to differentiate arteries and veins, particularly within the cavernous sinus and occasionally CTA may not give an accurate assessment of the position of the carotid arteries. Preoperative digital subtraction angiography may then be necessary and angiography should always be performed when CT raises the possibility of an aneurysm and MRA is contraindicated.

Nuclear isotope imaging techniques

Nuclear medicine techniques, such as positron emission tomography (PET) or single-photon emission tomography (SPECT), have been used to obtain in vivo characterization of tissue. The presence of octreotide-binding somatostatin receptors in nonfunctioning adenomas cause them to take up 111In-DTPA-octreotide but meningiomas may also express somatostatin receptors and take up somatostatin receptor-specific isotopes (7). However tracers that bind to the enzyme monoamino-oxidase β‎ have been used to differentiate meningioma from pituitary adenoma using PET and more recently a D2 dopamine receptor specific isotope [18F]fluoro-ethyl-spiperone has been reported to differentiate nonfunctioning adenomas from craniopharyngioma and meningioma (8). PET using tracers such as [18F]fluorodeoxyglucose and [11C]methionine can be used to study rates of glucose metabolism and protein synthesis. It can be used to assess tumours after treatment by differentiating viable tumour from scar tissue and monitoring pharmacological treatments. In the UK, the current availability of scanners limits the use of PET to selected patients and research.

Pituitary macroadenoma

Adenomas larger than 1 cm in maximum diameter are conventionally classified as macroadenomas irrespective of their endocrine characteristics. Imaging for diagnosis is performed to demonstrate the cause of an endocrine disturbance or for symptoms and signs of pituitary region pathology, e.g. visual loss. It is therefore directed at identifying the tumour and its extent. In cases of nonfunctioning adenoma the differential diagnosis includes other causes of pituitary region masses (see below). Once a macroadenoma is diagnosed the role of imaging is to localize the tumour for surgical or radiotherapy planning and to monitor the effects of therapy.

The imaging features of macroadenomas are generally similar for functioning and nonfunctioning pituitary tumours. They may extend well beyond the pituitary fossa but will cause expansion of the fossa as evidence of their origin (Fig. 2.3.5.4). On CT, solid tumours are isodense or hypodense relative to brain tissue and show variable patterns of enhancement after radiographic contrast media administration, on MRI signal return is typically similar to that of brain on both T1- and T2-weighted sequences (9). Gadolinium administration causes signal change due to shortening of the recovery time and brightening of tumour and gland. This is best demonstrated on T1-weighted sequences and usually normal gland enhances more avidly than tumour aiding its localization. Cysts or areas of necrosis cause foci of moderate hypointensity on T1-weighted and hyperintensity on T1-weighted sequences with heterogeneous enhancement after gadolinium administration.


Fig. 2.3.5.4 (a) Sagittal and (c) coronal T1-weighted and sagittal T1-weighted MR images showing a macroadenoma. The tumour had resulted in expansion of the sella and is extending into the chiasmatic cistern. The optic chiasm is so compressed that it cannot be seem. Within the solid tumour are small foci of bright signal on the T1-weighted sequences (a,c), which appear darker that the rest of the tumour on the T2-weighted sequence (b). These represent areas of necrosis and haemorrhage (arrows).

Fig. 2.3.5.4
(a) Sagittal and (c) coronal T1-weighted and sagittal T1-weighted MR images showing a macroadenoma. The tumour had resulted in expansion of the sella and is extending into the chiasmatic cistern. The optic chiasm is so compressed that it cannot be seem. Within the solid tumour are small foci of bright signal on the T1-weighted sequences (a,c), which appear darker that the rest of the tumour on the T2-weighted sequence (b). These represent areas of necrosis and haemorrhage (arrows).

Signal due to haemorrhage may produce a more specific pattern but these are complex because the magnetic effects of iron in haemoglobin, changes as the molecule degrades after red blood cell lysis. In general, these are best appreciated on T1-weighted MRI since within days of haemorrhage increased concentrations of methaemoglobin within areas of haemorrhage causes T1 recovery time shortening and bright signal on this sequence. The appearance is therefore similar to that seen after gadolinium administration and may only be recognized if unenhanced imaging is performed. In the acute period after haemorrhage (<3 days) MR changes are nonspecific but acute haemorrhage is hyperdense on CT. Thereafter methaemoglobin forms which has a characteristic signal and can persist for weeks so that subacute and chronic haemorrhage is more easily identified on T1-weighted MRI. It is due to shortening of the T1 signal recovery time and termed a paramagnetic effect. It is similar to the signal returned by fat and some tumour products secreted by lesions such as craniopharyngiomas. Physiological bright or hyperintense signal on T1-weighted sequences can be seen in the posterior lobe and stalk, as described above. This property of phospholipids is also described as paramagnetic and is due to similar shortening of T1 recovery time. Less intense and smaller areas of bright signal are not uncommon in tumours of patients without a history suggestive of apoplexy. These asymptomatic haemorrhages have been confirmed surgically in a proportion of patients. The scan appearances of tumours in patients presenting acutely with pituitary apoplexy will reflect the relative extent of haemorrhage or necrosis, with gadolinium enhancement evident at the margins of necrotic areas (10) (see Fig. 2.3.5.4).

There have been attempts to correlate tumour appearances on imaging with hormonal activity. Imaging features such as tumour size, evidence of local invasion, CT density, and MR signal have been correlated with hormone production (9). Correlations have thus been based on general imaging features such as extreme size in gonadotroph adenomas and the presence of hypointense foci (probably due to haemorrhage or necrosis) on T2-weighted scans in growth hormone-secreting tumours. But MR, although capable of measuring fundamental chemical characteristics does not give a hormone-specific image (11). Its contribution to patient management is currently to accurately delineate tumour extent and effect on adjacent structures. The preoperative MRI appearances are helpful in directing the surgeon to likely areas of local invasion. Administration of gadolinium aids the identification of normal pituitary gland from tumour. Tumour invasion of the cavernous sinus, sphenoid bone, and extension into the chiasmatic cistern are evident on MRI. Such behaviour has been identified surgically and histologically in all tumour types. Scotti et al. studied the MRI features useful in the preoperative diagnosis of cavernous sinus invasion. Encasement of the carotid artery was the most specific sign of cavernous sinus invasion. Asymmetry of the cavernous sinuses, displacement of the lateral wall and of the carotid artery were inconsistent features of invasion. An indistinct medial sinus wall was an unreliable feature, being common in controls. Intraoperative MRI is being developed for determining the completeness of transsphenoidal resections but because magnets of low field strengths have to be used, image quality remains a concern (12).

After surgery, the timing of follow-up scans is important since early postsurgical changes due to local swelling (in the first 1–2 weeks) and surgical packing materials, used in the transsphenoidal exposure, may be confused with remnant tumour (13). Gelfoam packing material returns hypointense or hyperintense signal and enhances on early postoperative MRI after gadolinium. Its reabsorption takes 4–15 months. Biological packing material returns mixed signal with fat being hyperintense and muscle isointense on T1-weighted MRI. Re-alignment of the normal pituitary gland and reabsorption of packing material is usually evident on follow-up scans at 3 months. Early scanning is therefore only useful to investigate possible surgical complications and scanning to identify residual tumour is best delayed for at least 3 months.

Demonstration of tumour regression or recurrence relies on comparisons between follow-up scans. The protocols for post-operative follow-up imaging involves a baseline study obtained 3 months after hypophysectomy followed by interval MRI according to tumour type (usually annually for 5 years). More frequent scans are obtained if the patient’s visual fields change or histological examination of the resected tumour suggests local invasion. Patients treated medically for functional macroadenomas are also monitored by serial imaging, in combination with biochemical and clinical follow-up assessments. A reduction in the size of prolactin-secreting macroadenomas after treatment with dopamine agonists may be accompanied by haemorrhage and therefore changes in signal returned by the tumour on MRI (14). Growth hormone-secreting tumours may also shrink in response to treatment with somatostatin analogues but this is less consistent and changes of necrosis or haemorrhage are infrequent. Reductions in tumour size can be demonstrated within weeks but continued shrinkage has been documented up to 3 years after starting treatment. Early follow-up imaging is therefore useful to document tumour response. A baseline study is obtained when the patient starts treatment and is then repeated 3 and 12 months later. Subsequent imaging is performed in regard to the response to therapy and dictated by the clinical and biochemical examinations.

Microadenomas

The demonstration of pituitary microadenoma remains a major diagnostic challenge for pituitary imaging. To identify adenomas less than 10 mm in size demands the highest standards of technique and interpretation. In most patients the presence of a microadenoma is assumed from biochemical testing and imaging is undertaken to confirm an intrasellar source and to guide its transsphenoidal excision. MRI is superior to CT for both diagnosis and localization (15) of microadenomas, since they show little inherent contrast to normal pituitary tissue on CT and scanning require injection of intravenous radiographic contrast agents to demonstrate nonenhancement of the microadenoma against a background of normal gland enhancement (Fig. 2.3.5.5). On MRI, microadenomas are typically spherical or oval in shape and return signal hypointense relative to normal anterior lobe on T1-weighted and hyperintense on T2-weighted sequences. Prolactinomas usually appear bright on T2-weighted sequences, whereas growth hormone-secreting tumours are more likely to be isointense or hypointense on T2-weighted sequences (16).


Fig. 2.3.5.5 Coronal (a) T1-weighted and (b) T1-weighted gadolinium-enhanced MR images showing marked enhancement of the normal sized gland after gadolinium administration. There is a microadenoma (arrows) in the right side of the gland. This enhances less avidly than the gland and so appears less bright.

Fig. 2.3.5.5
Coronal (a) T1-weighted and (b) T1-weighted gadolinium-enhanced MR images showing marked enhancement of the normal sized gland after gadolinium administration. There is a microadenoma (arrows) in the right side of the gland. This enhances less avidly than the gland and so appears less bright.

The need for high precision scanning has stimulated research to develop better MRI techniques. Most centres perform scans in the coronal and sagittal planes, initially without gadolinium enhancement. The coronal plane is best and typical parameters for a T1-weighted spin echo sequence are TR 500 ms, TE 25 ms, 3 mm contiguous slice thickness with four excitations, which requires 8–9 min scan time. A 3D volume scan has the theoretical advantage of allowing postprocessing of images in different planes, higher resolutions and better SNRs but adds to the overall scan time. Techniques such as 3D-SPGR and 3D-FLASH have demonstrated the utility of this approach. The value of T1-weighted sequences with IV gadolinium enhancement is limited by the relatively poor uptake of gadolinium by adenomas. So after gadolinium administration, microadenomas are usually evident as hypointense lesions within an enhancing gland. If the tumour is very small the gland enhancement may mask its relative hypointensity and use of half-dose (0.05 mmol/kg) gadolinium and delayed scanning techniques are performed to improve detection rates (16). Simply increasing the magnetic field strength and the SNR of the scan does not appear to solve the problem (17).

An alternative approach to improve microadenoma detection rates is dynamic MRI. This technique employs rapid sequential imaging to show temporal differences in gadolinium uptake between adenoma and normal gland. In this way, microadenomas which enhance later than the surrounding normal gland can be identified. Early studies showed that normal pituitary enhanced before adenomas, and using faster acquisition times (5–10 s per image) Yuh et al. found that macroadenoma enhanced at the same time as the posterior lobe and before the anterior lobe suggesting that they have a direct blood supply (18). Dynamic scanning with various fast imaging techniques have been used to identify microadenomas not detected on scanning with conventional enhancement protocols but there is an increased rate of false positives (19). The sensitivity of nonenhanced high-resolution MRI for pituitary microadenoma is in the order of 60–80%. Conventional scanning with contrast enhancement detects 5–10% more lesions and dynamic scanning a further 5–10% of lesions (20). Detection rate can thus be improved but at the expense of a higher rate of false-positive results. The problems associated with imaging at this level are both technical and biological. Technically dynamic MRI demands the maximum of humans and machines and both can be frustrated by the occurrence of small coincidental pituitary lesions, the so-called incidentalomas. Their frequency is difficult to gauge from the literature but Chong et al. found focal hypointensities in the pituitary glands of 38% of normal volunteers (21). That such ‘lesions’ exist, whatever their incidence, means that we are unlikely to ever achieve 100% specificity rates on imaging alone.

Patients with Cushing’s syndrome and negative imaging may be further investigated by the venous effluent of the pituitary to differentiate Cushing’s disease from ectopic sources of adrenocorticotropic hormone (22). Simultaneous bilateral sampling after stimulation with corticotropin-releasing hormone is highly accurate but the test is invasive and carries a small risk of neurological complications. The technique is reserved for patients with normal pituitary imaging but the reader should appreciate that the extent to which imaging is pursued in order to exclude a microadenoma varies from centre to centre (19).

Other tumours of the suprasellar and parasellar regions

The preoperative differentiation of pituitary adenomas from other causes of sellar and parasellar tumours relies on imaging. The most common problem in practice, is to distinguish nonfunctioning pituitary macroadenomas from craniopharyngiomas, meningioma, and rarer causes of tumour in this region. Clinical symptoms and signs are usually unhelpful, with the exceptions of diabetes insipidus, which suggests craniopharyngioma and precocious puberty which suggests a primary hypothalamic lesion. There is a wide gamut of pathologies that may simulate nonfunctioning pituitary tumour and the position of mass lesions, relative to the optic chiasm, is a useful way of refining the differential diagnosis.

Lesions arising above optic chiasm

Lesions that arise above the chiasm include ependymoma, craniopharyngioma, haemangioblastoma, glioma (usually astrocytoma), hamartoma of the hypothalamus and lipoma. Ependymoma and haemangioblastoma and rarely craniopharyngioma may arise within the anterior part of the third ventricle (Fig. 2.3.5.6). Involvement of the hypothalamus by hamartomas, glioma, teratoma, or lipoma causes precocious puberty. Local pressure effects from optic chiasm tumour or an arachnoid cyst may also cause precocious puberty. In children with precocious puberty, hypothalamic hamartoma will be the cause in a third of cases though paradoxically larger hamartomas are less likely to cause this endocrine disturbance. Unlike other tumours in the region of the hypothalamus, hamartomas are isodense on CT and isointense on MRI relative to grey matter (see Fig. 2.3.5.6). They also, neither calcify nor enhance after administration of IV contrast media and thereby can usually be distinguished from craniopharyngioma and glioma.


Fig. 2.3.5.6 Sagittal T1-weighted MR image showing a nodular mass arising from the floor of the third ventricle. The lesion has the typical appearance of a hamartoma being isointense with brain.

Fig. 2.3.5.6
Sagittal T1-weighted MR image showing a nodular mass arising from the floor of the third ventricle. The lesion has the typical appearance of a hamartoma being isointense with brain.

Lesions arising below optic chiasm

The optic chiasm will be depressed by lesions arising in the floors of the third ventricle but elevated by suprasellar extension of intrasellar or parasellar tumours. The latter include meningioma, aneurysm, schwannoma (particularly of the trigeminal nerve), lymphoma, metastases, and tumours arising in bone. These lesions should be considered in the differential diagnosis of pituitary macroadenomas as well as the rare tumours of the neurohypophysis: pilocystic astrocytoma and granular cell tumour or choristoma. Imaging must distinguish intracranial aneurysm and the possibility of this diagnosis was, prior to MRI, an indication for preoperative intra-arterial angiography (IA-DSA). Blood flow in an aneurysm sac should be evident on MRI, which can be supplemented by MR or CT angiography. However if doubt remains, and rarely is this the case, intra-arterial angiography should be performed. Meningioma in this region may be parasellar and invade the cavernous sinus or arises from the tuberculum sellae. CT may show calcification and hyperostosis of bone but MRI is best at defining tumour extent. Meningiomas typically enhance homogeneously after gadolinium administration. Tumour arising in the sphenoid bone and clivus, such as chordoma, giant cell tumour and chondrosarcoma, or carcinomas arising in the nasopharynx, are associated with bone destruction, calcification, and a variable degree of enhancement. They may simulate bone invasion by pituitary macroadenoma. Imaging by both CT and MRI is helpful, since the former will demonstrate bone erosion and the latter the effects of the tumour on adjacent tissues (Fig. 2.3.5.7). Intrasellar or suprasellar metastases should always be considered in the differential diagnosis since their CT and MRI appearance is variable and they can be indistinguishable on imaging from other tumour types.


Fig. 2.3.5.7 Chordoma arising in the sphenoid bone. Sagittal (a) T1-weighted and (b) T1-weighted gadolinium-enhanced, and (c) axial CT showing a well-defined calcified tumour elevating and displacing the pituitary gland (arrows) anteriorly. Note how the gland enhances but there is no enhancement of the tumour.

Fig. 2.3.5.7
Chordoma arising in the sphenoid bone. Sagittal (a) T1-weighted and (b) T1-weighted gadolinium-enhanced, and (c) axial CT showing a well-defined calcified tumour elevating and displacing the pituitary gland (arrows) anteriorly. Note how the gland enhances but there is no enhancement of the tumour.

Lesions arising in the chiasmatic cistern

Finally, tumours may arise in the chiasmatic cistern and simulate suprasellar extension of a macroadenoma. The differential diagnosis includes: optic nerve glioma, meningioma, craniopharyngioma (Fig. 2.3.5.8), aneurysm, and metastasis. Again MRI has made a substantial contribution to refining the preoperative diagnosis in this region. Its key attribute lies in its ability to identify the anterior optic pathway and thereby distinguish optic nerve glioma from extra-axial tumour. This largely depends on using T1-weighted coronal images to follow the pathway from the optic nerves to the tracts. Sumida et al. found that the optic nerves, chiasm, and tracts could be visualized in over 84% of patients with pituitary adenoma, craniopharyngioma, or Rathke’s cleft cyst (23) (Fig. 2.3.5.9). The chiasm and tracts could be identified in 85% of 14 patients with meningioma but in only 50% were the optic nerves visible, reflecting the frequent anterior location of this tumour. Gadolinium-enhanced scanning is useful for the identification of meningioma and metastases. Optic nerve glioma involving the chiasm rarely enhances and the imaging diagnosis depends on identifying enlargement of the optic pathway as spread is transneural. Other features of neurofibromatosis type 1 are evident in a quarter of patients with optic nerve gliomas, so imaging should include the whole cranium. Optic nerve and chiasm gliomas are isointense to grey matter on T1-weighted and isointense or hyperintense on T2-weighted sequences. Involvement of the optic tracts and brain parenchyma is identified as hyperintense signal on T1-weighted sequences. Unless tumours are very large it is usually possible to see CSF between the inferior tumour margin from the diaphragma sellae and so to exclude suprasellar extension of an intrasellar mass.


Fig. 2.3.5.8 Craniopharyngioma: (a) Coronal T1-weighted, (b) axial CT, (c) coronal T1-weighted gadolinium-enhanced, and (d) sagittal MR image. The imaging appearances of craniopharyngioma depend on how much of the tumour is cystic and how much solid. Typically tumours have a relatively circumscribed outline; cystic areas do not enhance but they may be septated with thick enhancing walls whereas solid elements do enhance. Calcification patterns vary from solid lumps to popcorn-like foci or less commonly an eggshell pattern lining the wall of a cyst. In addition to calcification these tumours may contain paramagnetic substances which, like gadolinium enhancement, appear bright on T1-weighted MR images. Craniopharyngioma cysts contain variously, cholesterol, triglycerides, methaemoglobin, and desquamated epithelium. High concentrations of protein or methaemoglobin affect the MR signal and produce this characteristic T1-weighted hyperintensity of a paramagnetic substance, as evident in one of the cystic components in (a). Imaging protocols therefore need to include both CT and MR scanning, since the former is the more sensitive means of detecting calcification. Craniopharyngiomas on CT are typically of mixed alternative with or without calcification. Contrast enhancement with gadolinium improves tumour definition from normal structures. The signal of solid tumour is isointense or hypointense relative to brain on precontrast T1-weighted sequences and enhances after gadolinium administration. On T1-weighted sequences it is usually of mixed hypo or hyperintensity. Larger areas of calcification are usually hypointense on both T1- and T2-weighted sequences. Cysts are hyperintense on T2-weighted and hypointense on T1-weighted sequences but if they contain paramagnetic substances, particularly breakdown products of haemoglobin, this pattern will be reversed or they may appear hyperintense on both T2-weighted and T1-weighted sequences.

Fig. 2.3.5.8
Craniopharyngioma: (a) Coronal T1-weighted, (b) axial CT, (c) coronal T1-weighted gadolinium-enhanced, and (d) sagittal MR image. The imaging appearances of craniopharyngioma depend on how much of the tumour is cystic and how much solid. Typically tumours have a relatively circumscribed outline; cystic areas do not enhance but they may be septated with thick enhancing walls whereas solid elements do enhance. Calcification patterns vary from solid lumps to popcorn-like foci or less commonly an eggshell pattern lining the wall of a cyst. In addition to calcification these tumours may contain paramagnetic substances which, like gadolinium enhancement, appear bright on T1-weighted MR images. Craniopharyngioma cysts contain variously, cholesterol, triglycerides, methaemoglobin, and desquamated epithelium. High concentrations of protein or methaemoglobin affect the MR signal and produce this characteristic T1-weighted hyperintensity of a paramagnetic substance, as evident in one of the cystic components in (a). Imaging protocols therefore need to include both CT and MR scanning, since the former is the more sensitive means of detecting calcification. Craniopharyngiomas on CT are typically of mixed alternative with or without calcification. Contrast enhancement with gadolinium improves tumour definition from normal structures. The signal of solid tumour is isointense or hypointense relative to brain on precontrast T1-weighted sequences and enhances after gadolinium administration. On T1-weighted sequences it is usually of mixed hypo or hyperintensity. Larger areas of calcification are usually hypointense on both T1- and T2-weighted sequences. Cysts are hyperintense on T2-weighted and hypointense on T1-weighted sequences but if they contain paramagnetic substances, particularly breakdown products of haemoglobin, this pattern will be reversed or they may appear hyperintense on both T2-weighted and T1-weighted sequences.


Fig. 2.3.5.9 Rathke’s cleft cyst: (a) sagittal T1-weighted and (b) coronal T2-weighted MR images showing a cyst in the right side of the sella. Symptomatic Rathke’s cleft cysts are pathological enlargements of remnant of Rathke’s pouch and typically arise between the two lobes of the pituitary. They are lined by cuboidal or columnar epithelium and contain a variety of fluid types with differing magnetic properties. On MRI, the signal returned is variable; it may follow that of cerebrospinal fluid (i.e. hypointense on T1-weighted and hyperintense on T2-weighted sequences) or have signal characteristic of a protein-rich fluid (i.e. hyperintense on both T1- and T2-weighted sequences) or that of altered haemorrhage (i.e. hyperintense on T1-weighted and hypointense on T2-weighted sequences). Although the cyst wall does not usually enhance, intravenous administration of gadolinium is useful in order to exclude craniopharyngioma and to show the position of the normal pituitary (24). The CT appearance is typically of a noncalcified intrasellar cystic mass but showing that a lesion is not calcified is helpful in distinguishing from a craniopharyngioma. Cysts in atypical locations may make differentiation from an arachnoid cyst of the chiasmatic cistern (which also do not enhance) difficult. Rarely, they are found in the nasopharynx associated with persistent remnants of the craniopharyngeal duct. Another atypical feature of these cysts is the occasional finding of a central hypointense or isointense focus within the cyst. This is presumed to be due to desquamatized epithelium forming a ‘waxy nodule’ (25).

Fig. 2.3.5.9
Rathke’s cleft cyst: (a) sagittal T1-weighted and (b) coronal T2-weighted MR images showing a cyst in the right side of the sella. Symptomatic Rathke’s cleft cysts are pathological enlargements of remnant of Rathke’s pouch and typically arise between the two lobes of the pituitary. They are lined by cuboidal or columnar epithelium and contain a variety of fluid types with differing magnetic properties. On MRI, the signal returned is variable; it may follow that of cerebrospinal fluid (i.e. hypointense on T1-weighted and hyperintense on T2-weighted sequences) or have signal characteristic of a protein-rich fluid (i.e. hyperintense on both T1- and T2-weighted sequences) or that of altered haemorrhage (i.e. hyperintense on T1-weighted and hypointense on T2-weighted sequences). Although the cyst wall does not usually enhance, intravenous administration of gadolinium is useful in order to exclude craniopharyngioma and to show the position of the normal pituitary (24). The CT appearance is typically of a noncalcified intrasellar cystic mass but showing that a lesion is not calcified is helpful in distinguishing from a craniopharyngioma. Cysts in atypical locations may make differentiation from an arachnoid cyst of the chiasmatic cistern (which also do not enhance) difficult. Rarely, they are found in the nasopharynx associated with persistent remnants of the craniopharyngeal duct. Another atypical feature of these cysts is the occasional finding of a central hypointense or isointense focus within the cyst. This is presumed to be due to desquamatized epithelium forming a ‘waxy nodule’ (25).


Fig. 2.3.5.10 Craniopharyngioma of adamantinous type. Craniopharyngioma arises from remnants of Rathke’s pouch. The peak age at diagnosis is 5–15 years with a second peak in late middle age (55–75 years). This bimodal age distribution of incidence has been linked to histological tumour variants: an adamantinous type being commoner in children and young adults and a squamous-papillary type in older patients (26). Approximately 50% of patients present under the age of 20 years and calcification is commoner in younger patients; being demonstrable in 70% of tumours. Tumour usually are found above the sella, with 5% purely intrasellar, 20% purely suprasellar, and 75% both intrasellar and extrasellar. There have been attempts to correlate the heterogeneous MR images with the two histological variants (27). The adamantinous types were either mixed solid-cystic or mainly cystic tumours with hyperintense cysts on T1-weighted MR images. The squamous-papillary types were solid or mixed solid-cystic tumours with hypointense cysts on T1-weighted MR images. Imaging features found to be discriminating were: encasement of vessels, a lobulated shape, and hyperintense cyst signal for the former and a round shape, hypointense cysts and a predominantly solid appearance for the latter. Calcification is commoner in the adamantinous tumour but not discriminatory.

Fig. 2.3.5.10
Craniopharyngioma of adamantinous type. Craniopharyngioma arises from remnants of Rathke’s pouch. The peak age at diagnosis is 5–15 years with a second peak in late middle age (55–75 years). This bimodal age distribution of incidence has been linked to histological tumour variants: an adamantinous type being commoner in children and young adults and a squamous-papillary type in older patients (26). Approximately 50% of patients present under the age of 20 years and calcification is commoner in younger patients; being demonstrable in 70% of tumours. Tumour usually are found above the sella, with 5% purely intrasellar, 20% purely suprasellar, and 75% both intrasellar and extrasellar. There have been attempts to correlate the heterogeneous MR images with the two histological variants (27). The adamantinous types were either mixed solid-cystic or mainly cystic tumours with hyperintense cysts on T1-weighted MR images. The squamous-papillary types were solid or mixed solid-cystic tumours with hypointense cysts on T1-weighted MR images. Imaging features found to be discriminating were: encasement of vessels, a lobulated shape, and hyperintense cyst signal for the former and a round shape, hypointense cysts and a predominantly solid appearance for the latter. Calcification is commoner in the adamantinous tumour but not discriminatory.

Inflammatory diseases of the pituitary chiasmatic cistern

Involvement of the pituitary region by sarcoidosis, Langerhans’ cell histiocytosis, tuberculous meningitis or abscesses may cause endocrine symptoms such as diabetes insipidus and abnormalities on imaging. In Langerhans’ cell histiocytosis (also called eosinophilic granulomatosis) patients present with diabetes insipidus and bone lesions. The former frequently leads to imaging of the hypothalamic pituitary axis and on MRI the normal posterior pituitary high signal is typically absent (7). The diagnosis may be made by recognition of bone lesions which often occur in the skull—one of the rare situations when plain skull radiographs may be helpful in diagnosis of pituitary region disease.

Granulomatous leptomeningitis is more frequently evident in the basal cisterns than elsewhere in the cranium and involvement of the chiasmatic cistern is more likely to be recognized because patients present with visual or endocrine symptoms. Sarcoid granulomas may be identified as meningeal masses isodense on CT and isointense on MRI, relative to grey matter (8). They are best demonstrated on T1-weighted MRI with gadolinium enhancement. Other foci of meningeal enhancement should be sought since it may be necessary to resort to biopsy to confirm the diagnosis and a more superficial focus would be more accessible. The differential for pathological meningeal enhancement in the suprasellar region includes metastatic tumour and it is important to keep this possibility in mind. In a report of two patients (28), initial imaging performed for idiopathic diabetes insipidus was negative, but 1–2 years later, metastatic germinoma was demonstrated in the hypothalamus. Interval imaging and a high degree of suspicion is therefore warranted.

Diffuse enhancement of the pituitary occurs in lymphocytic hypophysitis and other rare forms of hypophysitis on MRI (Fig. 2.3.5.11). The gland and in particular the anterior lobe are usually moderately enlarged so hypophysitis is difficult to distinguish from adenoma on imaging alone. This rare form of autoimmune endocrine disease has probably been under diagnosed in the past because it was mistaken for adenoma or unrecognized prior to MRI being more available (29).


Fig. 2.3.5.11 Axial (a) T1-weighted and (b) T1-weighted gadolinium-enhanced MR images showing an intrasellar abscess. The appearances are of a mass replacing the gland with prominent enhancement of the margins of the abscess and the stella meininges.

Fig. 2.3.5.11
Axial (a) T1-weighted and (b) T1-weighted gadolinium-enhanced MR images showing an intrasellar abscess. The appearances are of a mass replacing the gland with prominent enhancement of the margins of the abscess and the stella meininges.

References

1. Fox WC, Wawrzyniak S, Chandler WF. Intraoperative acquisition of three-dimensional imaging for frameless stereotactic guidance during transsphenoidal pituitary surgery using the Arcadis Orbic System. J Neurosurg, 2008; 108: 746–50.Find this resource:

2. Brooks BS, el Gammal T, Allison JD, Hoffman WH. Frequency and variation of the posterior pituitary bright signal on MR images. AJNR Am J Neuroradiol, 1989; 10: 943–8.Find this resource:

3. Kucharczyk W, Lenkinski RE, Kucharczyk J, Henkelman RM. The effect of phospholipid vesicles on the NMR relaxation of water: an explanation for the MR appearance of the neurohypophysis?. AJNR Am J Neuroradiol, 1990; 11: 693–700.Find this resource:

4. Scott WA. Magnetic Resonance Imaging of the Brain and Spine. Philadelphia, New York: Lippincott-Raven, 1996:59–63.Find this resource:

    5. Sakamoto Y, Takahashi M, Korogi Y, Bussaka H, Ushio Y. Normal and abnormal pituitary glands: gadopentetate dimeglumine-enhanced MR imaging. Radiology, 1991; 178: 441–5.Find this resource:

    6. Friedman TC, Zuckerbraun E, Lee ML, Kabil MS, Shahinian H. Dynamic pituitary MRI has high sensitivity and specificity for the diagnosis of mild Cushing’s syndrome and should be part of the initial workup. Horm Metab Res, 2007; 39: 451–6.Find this resource:

    7. Tien RD, Naston TH, McDermott MW, Dillon WP, Kucharczyk J. Thickened pituitary stalk on MR images in patients with diabetes insipidus and Langerhaus cell histiocytosis. AJNR Am J Neuroradiol, 1990; 11: 703–8.Find this resource:

      8. Engelken JD, Yuh WTC, Carter KD, Nerad JA. Optic nerve sarcoidosis. MR findings. AJNR Am J Neuroradiol, 1992; 13: 228–30.Find this resource:

      9. Davis PC, Hoffman JC Jr., Tindall, GT, Braun IF. CT-surgical correlation in pituitary adenomas: evaluation in 113 patients. AJNR Am J Neuroradiol, 1985; 6: 711–16.Find this resource:

        10. Semple PL, Jane JA, Lopes MB, Laws ER. Pituitary apoplexy: correlation between magnetic resonance imaging and histopathological results. J Neurosurg, 2008; 108: 909–15.Find this resource:

        11. Lundin P, Nyman R, Burman P, Lundberg PO, Muhr C. MRI of pituitary macroadenomas with reference to hormonal activity. Neuroradiology, 1992; 34: 43–51.Find this resource:

        12. Ahn JY, Jung JY, Kim J, Lee KS, Kim SH. How to overcome the limitations to determine the resection margin of pituitary tumours with low-field intra-operative MRI during trans-sphenoidal surgery: usefulness of Gadolinium-soaked cotton pledgets. Acta Neurochir (Wien), 2008; 150: 763–71.Find this resource:

        13. Dina TS, Feater SH, Laws ER, Davis DO. MR of the pituitary gland postsurgery: serial MR studies following transspheniodal resection. AJNR Am J Neuroradiol, 1993; 14: 763–9.Find this resource:

          14. Lundin P, Bergström K, Nyman R, Lundberg PO, Muhr C. Macroprolactinomas: serial MR imaging in long-term bromocriptine therapy. AJNR Am J Neuroradiol, 1992; 13: 1279–91.Find this resource:

            15. Johnson MR, Hoare RD, Cox T, Dawson JM, Maccabe JJ, Llewelyn DE, et al. The evaluation of patients with a suspected pituitary microadenoma: computer tomography compared to magnetic resonance imaging. Clin Endocrinol, 1992; 36: 335–8.Find this resource:

            16. Bonneville JF, Bonneville F, Cattin F. Magnetic resonance imaging of pituitary adenomas. Eur Radiol, 2005; 15: 543–8.Find this resource:

            17. Stadnik T, Stevenaert A, Beckers A, Luypaert R, Buisseret T, Osteaux M. Pituitary microadenomas: diagnosis with two- and three-dimensional MR imaging at 1.5T before and after injection of gadolinium. Radiology, 1990; 176: 419–28.Find this resource:

            18. Yuh WTC, Fisher DJ, Nguyen HD, Tali ET, Gao F, Simonson TM, et al. Sequential MR enhancement pattern in normal pituitary gland and pituitary adenoma. AJNR Am J Neuroradiol, 1994; 15: 101–8.Find this resource:

            19. Tabarin A, Laurent F, Catargi B, Olivier-Puel F, Lescene R, Berge J, et al. Comparative evaluation of conventional and dynamic magnetic resonance imaging of the pituitary gland for the diagnosis of Cushing’s disease. Clin Endocrinol, 1998; 49: 293–300.Find this resource:

            20. Elster AD. High-resolution, dynamic pituitary MR imaging: standard of care or academic pastime?. AJR Am J Roentgenol, 1994; 163: 680–2.Find this resource:

              21. Chong BW, Kucharczyk W, Singer W, George S. Pituitary gland MR: a comparative study of healthy volunteers and patients with microadenomas. AJNR Am J Neuroradiol, 1994; 15: 675–9.Find this resource:

              22. Oldfield EH, Doppman JL, Nieman LK, Chrousos GP, Miller DL, Katz DA, et al. Petrosal sinus sampling with and without corticotrophin-releasing hormone for differential diagnosis of Cushing’s syndrome. N Eng J Med, 1991; 325: 897–905.Find this resource:

                23. Sumida M, Arita K, Migita K, Iida K, Kurisu K, Uozumi T. Demonstration of the optic pathway in sellar/juxtasellar tumours with visual disturbance on MR imaging. Acta Neurochir(Wien), 1998; 140: 541–8.Find this resource:

                24. Krenning EP, Kwekkeboom DJ, Bakker WH, Breeman WA, Kooij PP, Oei HY, et al. Somatostatin receptor scintigraphy with 111In-DTPA-D-Pne and 123I-Tyr3-octeotride: the Rotterdam experience in more than 1000 patients. Eur J Nucl Med, 1993; 20: 716–31.Find this resource:

                25. Lucignani G, Losa M, Moresco RM, Del Sole A, Matarrese M, Bettinardi V, et al. Differentiation of clinically non-functioning pituitary adenomas from meningiomas and craniopharyngiomas by positron emission tomography with [18F]fluoro-ethyl-spiperone. Eur J Nucl Med, 1997; 24: 1149–55.Find this resource:

                26. Sumida M, Uozumi T, Mukada K, Arita. K, Kurisu K, Eguchi K, et al. Rathke cleft cysts: correlation of enhanced MR and surgical findings. AJNR Am J Neuroradiol, 1994; 15: 525–32.Find this resource:

                27. Kucharczyk W, Peck WW, Kelly WM, Norman D, Newton TH, et al. Rathke cleft cysts: CT, MR imaging and pathological features. Radiology, 1987; 165: 491–5.Find this resource:

                28. Appignani B, Landy H, Barnes P. MR in central idiopathic diabetes insipidus in children. AJNR Am J Neuroradiol, 1993; 14: 1407–10.Find this resource:

                  29. Rivera JA. Lymphocytic hypophysitis: disease spectrum and approach to diagnosis and therapy. Pituitary 2006; 9: 35–45.Find this resource:

                  30. Buchfelder M, Nistor R, Fahlbusch R, Huk WJ. The accuracy of CT and MR evaluation of the sella turcica for detection of adrenocorticotropic hormone-secreting adenomas in Cushing disease. AJNR Am J Neuroradiol, 1992; 14: 1183–90.Find this resource:

                    31. Adamson TE, Wiestler OD, Kleihues P, Yasargil MG. Correlation of clinical and pathological features in surgically treated craniopharyngiomas. J Neurosurg, 1990; 73: 12–17.Find this resource:

                    32. Sartoretti-Schefer S, Wichmann W, Aguzzi A, Valavanis A. MR differentiation of adamantinous and squamous-papillary craniopharyngiomas. AJNR Am J Neuroradiol, 1997; 18: 77–87.Find this resource: