Imaging for neuro-ophthalmic and orbital disease
Article Outline
Purpose
To provide an update on imaging of the brain and orbit for ophthalmologists.
Design
Literature review.
Methods
A systematic English-language medline search and summary of recent literature on imaging of brain and orbit was performed.
Results
Computed tomography and magnetic resonance (MR) scanning are the mainstays for the evaluation of most disorders involving the brain and orbit. Computed tomography angiography and magnetic resonance angiography are relatively newer applications that are useful for the evaluation of arterial and venous disorders. Special sequences such as fat suppression and fluid attenuation inversion recovery are useful techniques for specific ophthalmic indications. Diffusion weighted imaging and perfusion-weighted imaging are improving the evaluation of acute stroke. Functional MRI, positron emission tomography scanning and single photon emission computed tomography may provide useful information regarding brain or tumor metabolism. Magnetic resonance spectroscopy has expanded our knowledge of brain function. Newer imaging studies have improved our diagnostic abilities on many fronts, including new sequences, new applications of imaging studies, and functional imaging of brain.
Conclusion
New imaging techniques for brain and orbit have an increased potential for improving diagnostic yield. Accurate and timely communication with the neuroradiologist can optimize the prescription and interpretation of imaging in ophthalmology.
Imaging techniques for visualizing pathology of the brain and orbit continue to evolve and improve. The clinician now has a wide variety of diagnostic tests from which to choose. This article provides a brief summary of the most commonly used techniques of interest to ophthalmologists1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 (e.g., computed tomography [CT] and magnetic resonance imaging [MR], digital subtraction angiography [DSA]); reviews the recent literature on newer modalities for imaging (e.g., MR angiography [MRA], CT angiography [CTA], MR venography, and diffusion-weighted imaging [DWI]);16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 updates clinicians regarding special MR imaging sequences (for example, postcontrast, fat suppression, fluid attenuation inversion recovery);31, 32, 33, 34, 35 and discusses functional imaging (e.g., functional MRI, positron emission tomography, and single photon emission computed tomography [SPECT]).36, 37, 38, 39 We emphasize the need to match the best imaging study to the clinical findings as well as underscore the importance of accurate and timely communication with the neuroradiologist.
Design
This study was a literature review.
Methods
An English-language Medline search was performed from 1996 to 2004 using the search terms, “neuroimaging”, “imaging”, “computed tomography”, “magnetic resonance imaging”, “angiography”, “diffusion-weighted imaging”, “magnetic resonance spectroscopy”, and “neuro-ophthalmology” and “orbit”. A content expert (A.G.L.) selected relevant titles and reviewed pertinent papers. Letters to the editor, case reports, and publications before 2000 were included only if they added significant or new information. A comprehensive update and review of the literature was used to provide a clinical summary of the available radiographic imaging techniques in ophthalmology.
Results
The most common indications for an ophthalmologist to order a neuroimaging study are as follows: (1) visual loss (for example, amaurosis fugax, optic neuropathy, junctional scotoma, bitemporal hemianopsia, and homonymous hemianopsia); (2) anisocoria or ptosis (e.g., Horner's syndrome, third nerve palsy); (3) proptosis (e.g., thyroid eye disease, orbital tumor or orbital pseudotumor, and carotid-cavernous fistula); (4) diplopia or ophthalmoplegia (e.g., ocular motor cranial neuropathy, supranuclear and internuclear ophthalmoplegia, gaze palsies, and orbital lesions); (5) oscillopsia (e.g., nystagmus); and (6) ophthalmoscopic abnormalities (e.g., papilledema, optic disk drusen,2 choroidal folds). We have chosen not to include the neuroimaging evaluation of isolated headache or facial pain in this manuscript, and the reader is referred to the several other excellent reviews from the neurologic literature.
The two main imaging techniques for the brain and orbit are CT and MR scans. The CT scan was first introduced in 1972 and is based on x-ray attenuation by tissues of various densities. Denser tissues block x-rays and the image is “brighter” on CT scan by convention. Orbital and cranial CT scans are widely available, easy to perform, relatively inexpensive and can be obtained rapidly. The contrast material for CT scanning is iodinated. Contrast allergy and renal failure may be contraindications to iodinated contrast for CT. Contrast improves the sensitivity and specificity of the CT scan and should be ordered in most cases. A CT scan performed for an acute hemorrhage, trauma (e.g., orbital fracture, traumatic optic neuropathy), thyroid eye disease [iodinated contrast may affect the treatment of systemic thyroid disease], or localization of an intraocular or intraorbital foreign body do not generally require contrast material because of the intrinsic contrast provided by the fat in the orbit. An orbital CT is different from a head CT, because it is angled differently and uses thinner slices (e.g., 3 mm vs 5 mm). It is often important to obtain at least two views of the orbit on imaging. The most commonly performed views are axial and coronal. Coronal and sometimes sagittal views can be obtained either directly or with computer reconstruction. Reconstructed views may be the only option in a patient who is unable to extend or flex the neck for a direct view. New software permits high quality reconstructed views that can be comparable to direct CT views.
A CT scan is not as sensitive as MR scanning for the detection of pathology in the brain, and an MR scan is the procedure of choice for almost all neuro-ophthalmic indications (e.g., cavernous sinus, posterior fossa, and dural venous sinus abnormalities). A CT scan is superior to a MR scan however in cases that require visualization of bone (e.g., fracture, hyperostosis, sphenoid wing agenesis in neurofibromatosis-1, craniosynostosis, paranasal sinus disease, certain clival pathology, bone destruction or erosion); calcification (e.g., optic nerve head drusen, meningioma, retinoblastoma); or acute intracranial hemorrhage (e.g., subarachnoid hemorrhage). A CT scan is also the procedure of choice in patients requiring an emergent study (e.g., acute stroke, brain abscess, pituitary apoplexy, intracranial shunt malfunction) or in patients who cannot undergo a MR scan (e.g., severe claustrophobia, cochlear implant, ferromagnetic aneurysm clip, pacemaker, or metallic foreign body). Although the widespread availability of CT scan and the rapid scan time make it more useful than MR for patients requiring emergent imaging, an MR scan may still be required later for optimum evaluation of the pathology.
Newer CT scanners can obtain continuous data in a helical fashion. Helical CT allows the CT tube and the detector to continually rotate around the patient. The advantages over conventional CT include shorter examination times, reduction of motion artifacts and radiation exposure, and superior sensitivity for detecting a foreign body.
An emergent CT scan might be ordered by an ophthalmologist for severe acute headache (“worst headache of my life”), an acute bitemporal hemianopsia or diplopia (e.g., pituitary apoplexy), subarachnoid hemorrhage (e.g., ruptured intracranial aneurysm), or acute papilledema. Patients with a history of intracranial shunting (e.g., ventriculoperitoneal shunt) procedure may present to the ophthalmologist with neuro-ophthalmic findings (e.g., dorsal midbrain syndrome, visual loss, optic atrophy, or papilledema) suggesting shunt failure. A CT scan for ventriculomegaly in this setting may be indicated. As the absence of hydrocephalus on neuro-imaging does not exclude shunt malfunction shunt series, or a nuclear medicine shuntogram may be necessary in some of these cases.35
Magnetic resonance imaging is based on signal detection of the interaction of protons within a powerful magnetic field (preferably 1.0 to 3.0 Tesla magnet) and thus, unlike CT, there are no x-rays for MR imaging. The physics behind magnetic resonance imaging (MRI) is beyond the scope of this manuscript, and the reader is referred to other more in-depth publications for further information on the history and physics of MRI. The MR study can be “weighted” toward T1 (better demonstrates normal anatomy) or T2 (better distinguishes pathology) to enhance the contrast between tissues of different signal intensities. For the ophthalmologist, the imaging appearance of fat (bright on T1) and cerebrospinal fluid (dark on T1 and bright on T2) might be helpful in reviewing MR studies. The intrinsic MR signal characteristics of melanin and hemorrhage might also be useful information for the ophthalmologist. Melanin is a paramagnetic substance that is bright on T1, and a subretinal hemorrhage might be distinguishable from a choroidal melanoma based on MRI. The ophthalmologist typically does not have to worry about ordering a T1- and T2-weighted study, as these sequences are routine for brain and orbit MR imaging.3, 4, 5, 6
The clinician should however be aware of some MR sequences that have specific applications in ophthalmology. Special MR sequences can suppress the normal bright signal of fat (Figure 1) on T1 (“fat suppression”) and bright cerebrospinal fluid on T2 (fluid attenuation inversion recovery or FLAIR). The FLAIR (Figure 2) sequences allow visualization of underlying pathology (e.g., white matter demyelination, posterior reversible encephalopathy) that might be obscured by the bright signal of normal cerebrospinal fluid. Likewise, fat suppression (Figure 1) allows visualization of abnormal bright signal on T1 without obscuration from the normal fat signal. Fat suppression should be ordered in orbital MR scans. Ophthalmologists however should be aware that incomplete fat suppression (caused by braces, dental hardware, metallic mascara, and air-bone interface of adjacent paranasal sinuses) can produce an artifact that might be misinterpreted as abnormal enhancement.32, 33, 34, 35 Gradient echo sequences can show hemorrhage on MR scans in patients with underlying vascular malformations, intracerebral hemorrhage, or traumatic brain injury. Although gradient echo imaging (in combination with FLAIR and T2 imaging) in MR can show subarachnoid hemorrhage, noncontrast CT is probably still the most commonly used study for subarachnoid hemorrhage because of the rapidity of CT, the potential artifacts in MR, and the potential for motion artifact in agitated patients.
FIGURE 1.
Orbital fat suppressed and nonfat suppressed magnetic resonance (MR) imaging. Orbital postcontrast T1 weighted MR imaging without (Top) and with fat suppression (Bottom). There is enhancement of the left optic nerve. The normal bright fat signal on T1-weighted imaging is suppressed using fat suppression.
FIGURE 2.
Fluid attenuation inversion recovery sequence. Axial brain T2 weighted MRI with fluid attenuation inversion recovery (FLAIR) shows periventricular white matter lesions. The normal bright cerebrospinal fluid signal is suppressed on FLAIR.
As with CT scanning, contrast material for MR scanning improves detection of underlying pathology by demonstrating areas of breakdown of the blood-brain barrier. Unlike the iodinated contrast for CT, the contrast material for MR is a paramagnetic material called gadolinium.7 Gadolinium contrast should be ordered in virtually all MR scans that are performed for a neuro-ophthalmic indication. Gadolinium is very safe and does not produce cross-reactions in patients with allergies to iodinated contrast or fluorescein dye. Magnetic resonance with gadolinium is also the procedure of choice in patients with renal insufficiency who require a contrast study.7
Wolintz and associates have emphasized the most common errors in ordering imaging for neuro-ophthalmic diagnoses.35 They divided the errors into prescriptive errors: “(1) failure to apply a dedicated study, (2) inappropriate use of a dedicated study, (3) omission of intravenous contrast, and (4) omission of specialized sequences); and interpretive errors including: (1) failure to detect the lesion because of misleading clinical information, (2) rejection of a clinical diagnosis because an expected imaging abnormality was absent, (3) assumption that a striking imaging abnormality accounted for the clinical abnormality, and (4) failure to consider the lack of clinical specificity of imaging abnormalities”. Most of the errors can be avoided by directly communicating with the radiologist before the ordering of the test and at the time of the interpretation of the film. Table 1 lists the common indications for ordering an imaging study. Table 2 summarizes some general guidelines for ordering and interpreting imaging in ophthalmology. Table 3 lists the imaging indications for third nerve palsy, and Table 4 lists the imaging indications for orbital disease.
TABLE 1. Neuro-ophthalmic Indications and Recommended Imaging Study
| Indication | Imaging Study | Contrast | Comment |
|---|---|---|---|
| Optic nerve drusen | CT scan of the orbit may show calcification | Not necessary | Orbital ultrasound is less costly and more sensitive than CT scan for drusen.2 |
| Papilledema | MRI head (with magnetic resonance venography) | Yes | Consider contrast magnetic resonance venography to exclude venous sinus thrombosis, especially in atypical cases of pseudotumor cerebri who are thin, male, or elderly. |
| Transient visual loss (amaurosis fugax) | MRA or CTA of neck for carotid stenosis or dissection. | Depends on clinical situation | May require adjunctive carotid Doppler study or catheter angiography. |
| Demyelinating optic neuritis | MRI head and orbit | Yes | FLAIR to look for demyelinating white matter lesions. MRI has prognostic significance for development of multiple sclerosis. |
| Inflammatory, infiltrative, or compressive optic neuropathy | MRI head and orbit | Yes | Fat suppression to exclude intraorbital optic nerve enhancement. CT is superior in traumatic optic neuropathy for canal fractures. |
| Junctional scotoma (i.e., optic neuropathy in one eye and superotemporal field loss in fellow eye) | MRI head (attention to sella) | Yes | |
| Bitemporal hemianopsia | MRI head (attention to chiasm and sella) | Yes | Consider CT of sella if an emergent scan is needed (e.g., pituitary or chiasmal apoplexy) or if imaging for calcification (e.g., meningioma or craniopharyngioma or aneurysm). |
| Homonymous hemianopsia | MRI head | Yes | Retrochiasmal pathway. DWI may be useful if acute ischemic infarct. If structural imaging negative and organic loss consider functional imaging |
| Cortical visual loss or visual association cortex (e.g., cerebral achromatopsia, alexia, prosopagnosia, simultagnosia, optic ataxia, Balint's syndrome) | MRI head | Yes | Retrochiasmal pathway. Consider DWI in ischemic infarct. If structural imaging negative and organic loss consider functional imaging (e.g., PET, SPECT, MRS) |
| Third, fourth, sixth nerve palsy or cavernous sinus syndrome. | MRI head with attention to the skull base. Isolated vasculopathic cranial neuropathies may not require initial imaging. See Table 3for third nerve palsy evaluation. | Yes | Rim calcification in aneurysm, calcification in tumors, and hyperostosis may be better seen on CT. |
| Internuclear ophthalmoplegia (INO), supranuclear or nuclear gaze palsies, dorsal midbrain syndrome, skew deviation | MRI head (brainstem) | Yes | Rule out demyelinating or other brainstem lesion. Include a FLAIR sequence. |
| Nystagmus | MRI brainstem | Yes | Localize nystagmus. |
| Hemifacial spasm | MRI brainstem (with or without MRA) | Yes | Facial nerve compression at root exit zone. |
| Horner's syndrome: preganglionic | MRI head and neck to second thoracic vertebra (T2) in chest with neck MRA | Yes | Rule out lateral medullary infarct, apical lung neoplasm, carotid dissection, etc. |
| Horner's syndrome: post-ganglionic | MRI head and neck to level of superior cervical ganglion (C4 level) with MRA neck | Yes | Rule out carotid dissection. Isolated post-ganglionic lesions are often benign. |
TABLE 2. Guidelines for Ordering Imaging Studies in Ophthalmology
| Decide whether a CT or MR scan is indicated. In most cases a MR scan is superior to CT for neuro-ophthalmic indications. CT is superior to MR for calcification, bone, acute hemorrhage, if an emergent scan is needed, or if the patient cannot undergo a MR scan. |
| Decide if contrast is needed. In most cases, contrast material should be ordered for all studies. Contrast may not be necessary in acute hemorrhage, thyroid ophthalmopathy, or trauma cases. |
| Localize the lesion clinically (“where is the lesion”) then order a study tailored to the location (e.g., head, orbit, neck). Take the time to fill out the radiographic order form personally with sufficient clinical detail for the radiologist to obtain the correct study. Don't just order a “brain MRI” for every case. |
| Consider ordering special imaging sequences (e.g., fat suppression for orbital post contrast study, fluid attenuation inversion recovery for white matter lesions, gradient echo for hemorrhage) depending on clinical indication. |
| Consider ordering special imaging for specific clinical indications (e.g., MRA or CTA, MRV, conventional arteriography). |
| Call the radiologist and tell him/her the differential diagnosis (“what is the lesion”) and the localization (“where is the lesion”). |
| If the imaging shows either no abnormality or an abnormality that does not match the clinical localization then call the radiologist or, better yet, review the films directly with him or her. Ask the radiologist if the area of interest has been adequately imaged, if artifact might be obscuring the lesion, or if additional studies might show the lesion. |
| If the clinical picture suggests a specific lesion or localization and initial imaging is “normal,” consider repeating the imaging with thinner slices and higher magnification of the area of interest, especially if the clinical signs and symptoms are progressive. |
| Recognize that the lack of an imaging abnormality does not exclude pathology. |
TABLE 3. Recommended Screening Neuroimaging Procedures for Patients with Acute Neurologically Isolated Third Cranial Nerve Palsy
| Clinical Characteristics | Study |
|---|---|
| Iris sphincter function normal; extraocular muscle function totally impaired (pupil-spared complete third nerve palsy) | None (in vasculopathic patient who improves). |
| Iris sphincter function totally impaired; extraocular muscle function impaired (pupil-blown third nerve palsy) | MRI (with MRA or CTA) followed by CA if MRI does not disclose a non-aneurysmal cause |
| Iris sphincter function partially impaired; extraocular muscle function totally impaired (relative pupil-sparing complete third nerve palsy) | MRI followed by MRA if MRI does not disclose a non-aneurysmal cause* |
| Iris sphincter function normal; extraocular muscle function partially impaired (pupil-sparing incomplete third nerve palsy) plus patient age > 40 years and vasculopathic risk factors present | MRI followed by MRA (or CTA) if MRI does not disclose a non-aneurysmal cause* |
TABLE 4. Orbital Indications and Recommendations for Imaging
| Orbital Lesion | Imaging Study | Contrast | Comment |
|---|---|---|---|
| Thyroid ophthalmopathy | CT or MR of orbit | Iodinated contrast may interfere with evaluation and treatment of systemic thyroid disease. | Bone anatomy is better seen on CT scan, especially if orbital decompression is being considered. |
| Orbital cellulitis and orbital disease secondary to sinus disease (e.g., silent sinus syndrome, sinusitis) | CT orbit and sinuses | Depends on clinical situation | MRI may be useful adjunct to CT scan, especially if possible concomitant cavernous sinus thrombosis is present. |
| Orbital inflammatory pseudotumor | CT or MR of orbit (with fat suppression) | Yes | Beware fat suppression artifact. |
| Orbital tumor (e.g., proptosis or enophthalmos, gaze- evoked visual loss) | CT or MR of orbit | Yes | Include head imaging if lesion could extend intracranially (e.g., optic nerve sheath meningioma). CT scan may be superior if looking for hyperostosis or calcification (e.g., sheath meningioma) |
| Orbital trauma (e.g., fracture, subperiosteal hematoma, orbital foreign body, orbital emphysema) | CT scan orbit with direct coronal | Not generally necessary | CT is superior to MR for bone fractures |
| Traumatic optic neuropathy | CT optic canal (thin sections) | Not generally necessary | CT is superior for visualizing fracture or bone fragment. MR may be superior for optic nerve sheath hemorrhage. |
| Carotid cavernous sinus or dural fistula (e.g., orbital bruit, arterialization of conjunctival and episcleral vessels, glaucoma) | CT or MR of head and orbit (with contrast-enhanced MRA) | Yes | CT or MR may show enlarged superior ophthalmic vein. May require catheter angiogram for final diagnosis and therapy. Color flow Doppler studies may be useful for detecting reversal of orbital venous flow. |
Although conventional CT and spin echo MR are the mainstays of imaging for the orbit and head, the ophthalmologist should be aware of some newer imaging techniques. Quantitative structural MR techniques include magnetization transfer imaging and diffusion-weighted imaging. Magnetization transfer MRI is based on the interaction among protons in environments, which have relatively free motion, and those that are more tightly bound (e.g., protons in macromolecules). Special techniques can saturate the magnetization of less mobile protons and reduce signal from observable magnetization. A low magnetization transfer ratio indicates a reduced ability of central nervous system macromolecules to exchange magnetization with surrounding water molecules. This might reflect damage to myelin or axonal membranes (e.g., multiple sclerosis).
Diffusion-weighted imaging is based on the microscopic random (Brownian motion) translational motion of water molecules (Figure 3). Changes in water molecular diffusion can be measured in vivo with DWI. This measurement of the self-diffusion coefficient of water indicates the mobility of water within tissue and is called the apparent diffusion coefficient. In DWI, a pair of pulsed magnetic field gradients are applied and water molecules that have diffused during the time interval between the applied pulses show a larger signal loss than water with restricted motion. Restricted diffusion appears bright on DWI. For an ophthalmologist, the most useful application of DWI is the ability to detect hyperacute ischemic lesions (bright on DWI) producing a homonymous hemianopsia or cortical visual impairment. Evolving DWI applications include the evaluation of inflammatory (e.g., brain abscess), degenerative (e.g., Alzheimer's dementia), demyelinating (e.g., multiple sclerosis), and neoplastic lesions (e.g., differentiating epidermoid vs arachnoid cyst).24, 25, 26, 27, 28, 29 Perfusion weighted imaging in combination with DWI can provide information about stroke pathophysiology and might help in decision making for utilizing interventional therapy in acute ischemic strokes (e.g., thrombolytic therapy).24, 25, 26, 27, 28, 29, 30, 31
FIGURE 3.
Diffusion-weighted imaging (DWI). This image shows acute ischemic infarcts that are bright on axial DWI.
In addition to conventional MR, other special MR techniques should be considered for specific ophthalmic indications. Although conventional catheter angiography remains the “gold standard” for vascular lesions (e.g., aneurysm, arteriovenous malformation, carotid-cavernous fistula), newer MR and CT angiography techniques are emerging that can reduce the need for diagnostic catheter angiography. Digital subtraction angiography is a conventional catheter angiographic method that shows contrast-filled vessels (e.g., arteries and veins) without interference from the background. An image of the brain is taken before (the “mask image”) and after the injection of contrast. The two images can be overlaid and a digital subtraction image can be produced that only shows the contrast-filled vessels. Postprocessing of the images by the neuroradiologist (e.g., windowing and filtering) can be performed after the image is acquired. Unfortunately, although the morbidity and mortality of modern digital subtraction angiography has been reduced substantially, there is still a small (< 0.5%) risk of severe complication (e.g., ischemic stroke) from catheter angiography.
One “high stakes” encounter that an ophthalmologist would need to consider with catheter angiography is the evaluation of the third cranial nerve palsy. The reader is referred to several other papers discussing the detailed clinical evaluation of the third nerve palsy.16, 17
Magnetic resonance angiography and CTA can be used to detect aneurysms (See Figures 4 and 5). MRA takes advantage of the flow-sensitive nature of MR signal. The two basic types of MRA are time-of-flight (TOF) and phase contrast (PC). Both TOF and PC MRA can acquire the data using two-dimensional (2 diopters [D]) or three-dimensional (3 D) scans. The two-D scan acquires each data slice individually and produces contiguous or overlapping data sets, while three-D scanning acquires the data in a block of tissue volume. The two-D TOF MRA is more sensitive to slow flow. The three-D technique allows thinner slices and has higher signal-to-noise ratio. A contrast-enhanced three-D acquisition optimizes the MRA technique (for example, possible dural venous sinus disease, evaluation of carotid stenosis). In contrast to TOF, PC MRA uses gradients to induce phase shifts in flowing blood. Both PC and TOF MRA are postprocessed to produce an image that resembles a conventional catheter angiogram. In general, TOF MRA is a more robust technique for evaluating vascular abnormalities of interest to ophthalmologists. The radiologist should look at the source images for the MRA however when reading the study especially for aneurysm. Jacobson and Trobe in a prior article summarized the recommendations for MRA.17 In Table 3, we have assumed that CTA is a study equivalent to MRA for detection of aneurysm in third nerve palsy. Contrast enhanced MRA and magnetic resonance venography improves MR angiography techniques and decreases artifacts. The advantages of MRA over CTA are the lack of ionizing radiation exposure, a nonnephrotoxic contrast (e.g., gadolinium), increased signal-to-noise ratio, and easier postprocessing. The advantages of CTA over MRA are increased spatial resolution and a technically easier image to acquire. The disadvantages of CTA include bone artifact near the skull base, cavernous sinus and orbital apex and the need for iodinated contrast. The sensitivity and specificity of the techniques is probably equivalent, although some authors have suggested that CTA may be superior to MRA for some indications. For example, multi-slice helical CT is faster than MRA.
FIGURE 4.
Magnetic resonance angiography. Magnetic resonance angiography (Top) shows a posterior communicating artery aneurysm in a patient with a third nerve palsy that is also seen on the conventional catheter angiogram (Bottom).
FIGURE 5.
Computed tomography angiography (CTA). There is a posterior communicating artery aneurysm seen on this CTA.
A MR scan is a superior study for assessing nonaneurysmal causes of a third nerve palsy, and therefore from a practical standpoint the combination of MRI with MRA instead of MR scan with CTA is likely to be the first line imaging for a patient with a third nerve palsy.16 Although the sensitivity for both MRA and CTA for the detection of an aneurysm producing a third nerve palsy is high (up to 98%), catheter angiography may still be required for cases with a pretest likelihood for harboring an aneurysm even with a negative MRA or CTA.17, 18, 19 It is likely that an aneurysm producing a third nerve palsy would be of sufficient size (≥5 mm) to be detected on MRA and CTA, but the “gold standard” remains catheter angiography for high suspicion cases.
Magnetic resonance angiography and CTA are also emerging as useful studies in the evaluation of vascular malformations, dissection, stenosis, or occlusion in hemispheric transient ischemic attacks, amaurosis fugax, or completed strokes. The major ophthalmic indication for performing a MR venogram is excluding dural venous sinus thrombosis in patients with increased intracranial pressure and papilledema. Contrast enhanced MRV reduces the incidence of artifacts seen with noncontrast MRV that might be relatively common in patients with pseudotumor cerebri.20, 21, 22
Magnetic resonance spectroscopy (MRS) is based on detecting various proton MR spectra (Figure 6).23 The four major resonances for MRS are: (1) choline-containing phoshoplipids, (2) creatine and phosphocreatine, (3) N-acetyl-aspartate (NAA), and (4) lactate. Reduction of NAA on MRS is a marker for neuronal loss. Certain tumors have no NAA (e.g., meningioma or metastases) or markedly decreased NAA (e.g., glioblastoma multiforme, metastasis). Creatine may be increased in hypometabolic states caused by ischemia or tumor (e.g., gliomatosis), or it may be decreased in hypermetabolic states. Choline is a component of cell membranes, and increased choline might suggest increased membrane synthesis in an active proliferating or a solid, hypercellular tumor. Creatine often remains stable in other disease processes and can be used as a control for MRS with levels of other metabolites expressed as a ratio to Cr (for example, increased CHO/Cr ratio in certain brain tumors). Other metabolites in brain can be detected with MRS including lipid and myoinositol, which are markers of gliosis and myelin damage. The normal brain derives energy from aerobic oxidation of glucose and does not normally have a significant lactate peak on MRS. Elevated lactate on MRS has been demonstrated in the occipital lobes of patients with metabolic disturbances, ischemia, trauma, and tumors. Magnetic resonance spectroscopy applications in neuro-ophthalmology include differentiating ischemic, neoplastic, demyelinating, radiation necrosis, inflammatory, and mitochondrial disorders affecting the visual pathway. Magnetic resonance spectroscopy has also been used anecdotally in evaluating intraocular tumors.
FIGURE 6.
Magnetic resonance spectroscopy (MRS). This MRS shows the magnetic resonance spectra for choline, creatinine, and N-acetyl-aspartate peaks.
Functional imaging studies that show physiology and metabolic function are useful in cases where structural imaging might be normal. Functional MR imaging is based on changes in T2 signal because of deoxyhemoglobin. This signal change is termed the blood oxygenation level-dependent mechanism. A local increase in neuronal activity leads to increased cerebral blood flow and a decrease in deoxyhemoglobin. The functional magnetic resonance imaging techniques have been used in some centers for presurgical mapping of brain function and have added new information regarding localization of visual function in the brain. Other functional imaging studies include positron emission tomography and single photon emission computed tomography. Positron emission tomography and SPECT both use radiolabeled molecules to image metabolism including regional blood flow and glucose metabolism. Positron emission tomography has superior sensitivity and tissue resolution than SPECT but is more expensive and less widely available. A positron emission tomography scan can image the whole body or single organs. The emerging and current applications for both functional studies include toxicity (e.g., cyclosporine-related cerebral blindness) neoplasms (e.g., metastatic disease), inflammatory disease (e.g., sarcoid), radiation necrosis, stroke, degenerative disorders, epilepsy, movement disorders, and migraine. The major applications for the ophthalmologist include patients with organic homonymous hemianopsias (for example, carbon monoxide poisoning, visual variant Alzheimer's disease) but normal structural imaging (for example, normal MR scan). Patients with functional imaging might show hypometabolism (“cold”) or hypermetabolism (“hot”) in areas where structural imaging studies (CT or MR scan) were normal.36, 37, 38, 39
Discussion
Ophthalmologists should be aware that MR scanning is superior to CT scanning for most intracranial neuro-ophthalmic indications. A CT scan may be the procedure of choice for acute hemorrhage, emergent scans, calcification, bone pathology, trauma, and for those patients who cannot undergo a MR scan. Computed tomography scanning is probably most cost effective for the majority of orbital lesions, although evaluations of optic nerve or orbital apex lesions are probably better seen with MRI. To optimize resources and maximize the diagnostic yield, the ophthalmologist should provide the radiologist with the suspected differential diagnosis and the localization of the presumed lesion to obtain the most effective neuroimaging study. Contrast material should usually be included in the study unless there is a contraindication. Vascular lesions, especially aneurysms, may be imaged using MRA and CTA but the catheter angiogram remains the “gold standard” in most cases. Special sequences such as fat suppression, FLAIR, and DWI are helpful in specific conditions. Functional imaging may be indicated in patients with normal structural imaging studies. Ophthalmologists should keep up to date on the newer techniques and applications of these imaging studies.
References
- . The role of medical imaging in the practice of neuro-ophthalmology . Radiol Clin North Am . 1987;25:863–876
- . A comparison of imaging techniques for diagnosing drusen of the optic nerve head . Arch Ophthalmol . 1999;117:1045–1049
- . MR imaging anatomy of the optic pathways . Radiol Clin North Am . 1999;37:1–36
- . Clinical signs and symptoms requiring computed tomography and magnetic resonance imaging evaluation . Neuroimaging Clin N Am . 1996;6:1–14
- . Neuroradiologic investigation of the visual system using magnetic resonance imaging . J Clin Neurophysiol . 1994;11:500–518
- . Neuroimaging in neuro-ophthalmology . Ophthalmology . 1989;96:555–568
- . Neuroimaging contrast agents in ophthalmology . Surv Ophthalmol . 2000;45:237–253
- Atypical features prompting neuroimaging in acute optic neuropathy in adults . Can J Ophthalmol . 2000;35:325–330
- . Neuroimaging techniques for the ophthalmologist . Can J Ophthalmol . 1997;32:277–281
- . Diagnostic tests in neuro-ophthalmology . Neurol Clin . 1996;14:201–222
- . Neuroimaging in neuro-ophthalmology . Neurol Clin . 1983;1:831–857
- . Application of new MR techniques in pediatric patients . Magn Reson Imaging Clin N Am . 2003;11:493–522
- . Newer MR imaging techniques for head and neck . Magn Reson Imaging Clin N Am . 2003;11:449–469
- Enhancement and demyelination of the intraorbital optic nerve (Fat suppression magnetic resonance imaging) . Ophthalmology . 1992;99:713–719
- . The localizing value of a quadrantanopia . Arch Neurol . 1997;54:401–404
- . The evaluation of isolated third nerve palsy revisited: an update on the evolving role of magnetic resonance, computed tomography, and catheter angiography . Surv Ophthalmol . 2002;47:137–157
- . The emerging role of magnetic resonance angiography in the management of patients with third cranial nerve palsy . Am J Ophthalmol . 1999;128:94–96
- . New techniques and applications for magnetic resonance angiography . Mt Sinai J Med . 2003;70:375–385
- . Magnetic resonance angiography . Adv Neurol . 2003;92:43–52
- . Manometry combined with cervical puncture in idiopathic intracranial hypertension . Neurology . 2002;58:26–30
- . Diagnostic criteria for idiopathic intracranial hypertension . Neurology . 2002;59:1492–1495
- . Isolated intracranial hypertension as the only sign of cerebral venous thrombosis . Neurology . 1999;53:1537–1542
- . MR spectroscopy of brain tumors . Magn Reson Imaging Clin N Am . 2003;11:415–429
- . Imaging of brain tumors with diffusion-weighted and diffusion tensor MR imaging . Magn Reson Imaging Clin N Am . 2003;11:379–401
- Diffusion-weighted MR Imaging: clinical applications in neuroradiology . Radiol Med (Torino) . 2003;106:521–548
- Diffusion and perfusion MR imaging . Rays . 2003;28:29–43
- . Diffusion-weighted magnetic resonance imaging and the evaluation of cortical blindness in preeclampsia . Surv Ophthalmol . 2003;48:647–650
- Imaging tutorial: differential diagnosis of bright lesions on diffusion-weighted MR images . Radiographics . 2003;23:e7
- Distinguishing ischemic stroke from stroke-like lesions of MELAS using apparent diffusion coefficient mapping . J Neurol Sci . 2003;216:11–15
- . Magnetic resonance-based techniques for the study and management of multiple sclerosis . Br Med Bull . 2003;65:133–144
- . Basic principles of diffusion-weighted imaging . Eur J Radiol . 2003;45:169–184
- Fat suppression in MR imaging: techniques and pitfalls . Radiographics . 1999;19:373–382
- . MRI Artifact masquerading as orbital disease . Int Ophthalmol . 2001;24:343–347
- . Contrast-enhanced fat-suppression neuroimaging . Neuroimaging Clin N Am . 1994;4:153–173
- Common errors in the use of magnetic resonance imaging for neuro-ophthalmic diagnosis . Surv Ophthalmol . 2000;45:107–114
- . Physiologic imaging in “functional” visual loss . Surv Ophthalmol . 1996;40:395–399
- Nuclear magnetic resonance spectroscopy (Principles and applications in neuroophthalmology) . Int Ophthalmol . 1994;18:171–181
- Late onset cyclosporine-induced cerebral blindness with abnormal SPECT imaging in a patient undergoing unrelated bone marrow transplantation . Bone Marrow Transplant . 2000;26:105–108
- . Quantitative cerebral blood flow imaging in a patient with the Heidenhain variant of Creutzfeldt-Jakob disease . Clin Nucl Med . 2001;26:770–773
Andrew G. Lee, M.D. is a graduate of the University of Virginia Undergraduate and School of Medicine. He completed his ophthalmology residency and was the chief resident at Baylor College of Medicine in Houston, Texas in 1993. Following residency, Dr. Lee completed a fellowship in neuro-ophthalmology at the Wilmer Eye Institute and was a post-doctoral Fight for Sight fellow at the Johns Hopkins Hospital in Baltimore, Maryland from 1993-1994. He was formerly an Associate Professor at Baylor College of Medicine and Adjunct Associate Professor at the M.D. Anderson Cancer Center in Houston from 1994-2000. He has published over 200 scientific articles and has authored two textbooks in ophthalmology. He is on the editorial board of five journals including the American Journal of Ophthalmology. He received the American Academy of Ophthalmology Achievement Award and the AAO Secretariat Award. Dr. Lee is currently Professor of Ophthalmology, Neurology, and Neurosurgery at the University of Iowa Hospitals and Clinics.
PII: S0002-9394(04)00819-0
doi:10.1016/j.ajo.2004.06.069
© 2004 Elsevier Inc. All rights reserved.
