DIAGNOSTIC IMAGING
Feb 7, 2009
This chapter provides an overview of diagnostic imaging in otolaryngology and head and neck surgery. It presents the principles and current modalities used in regional evaluation of the head and neck. The indications, applications, and limitations of conventional screening methods and high-technology imaging are delineated for a variety of otolaryngologic disorders.
PRINCIPLES OF IMAGING
A diagnostic imaging study is the most common consultation that otolaryngologists and head and neck surgeons request for a patient. Because many of the supporting and deep structures of the head and neck are beyond direct topographic evaluation, even with fiberoptic endoscopes and telescopes, further anatomic and physiologic information must be obtained for the temporal bone, skull base, paranasal sinuses, soft tissues of the neck, larynx, and other structures by means of diagnostic imaging. This does not mean that every patient needs an imaging study. Some patients need only a sensitive and thoughtful examiner. Other patients need conventional plain radiographs to complete the assessment. Others, however, need high-technology imaging for an accurate diagnosis and optimal treatment planning.
The influence of professional development and the team approach to diagnostic imaging is growing. Neuroradiologists have long gravitated to the field of head and neck radiology, and interventional specialization has extended far beyond this discipline. Angiography provides a model in which diagnostic imaging has led to interventional imaging and the training of dedicated specialists. A radiologic examination must be treated as a consultation to derive optimal information for formulating the best working diagnosis. This is the essence of effective diagnostic imaging, and it is achieved through effective communication between otolaryngologist and radiologist and through mutual analysis of the clinical problem and radiologic evidence. Consultation involves a well-prepared requisition, a telephone call, or a copy of the referring letter or consultation report.
The consulting team members must establish the optimal working diagnosis. Treatments may vary, but there is only one diagnosis, and it must be qualitatively and quantitatively effective (Table 6.1). Qualitative aspects of diagnosis usually lie within the field of the pathologist. However, the diagnostic image can provide qualitative answers to questions about whether disease exists, whether there is a specific disease process, such as a fracture, or whether a group disorder, such as bone destruction or cyst formation, can be identified. Diagnostic images also contain answers to quantitative questions about the extent of disease in all three dimensions, about disease that crosses regional boundaries that affect management, and about whether the disease is a local manifestation, metastatic condition, or systemic disorder. Imaging can help to inform the clinician about the presence and nature of a specific anatomic or physiologic derangement.
The foregoing are the types of issues clinicians formulate for imagers and for which solutions can be provided. However, routine questions receive only routine responses. Fifty percent of diagnostic information can be lost from computed tomography (CT) and magnetic resonance imaging (MRI) if the studies are not tailored to provide specific answers to pathologic anatomic and physiologic questions.
IMAGING MODALITIES
Conventional Imaging
The modalities representative of conventional imaging are listed in Table 6.2. Regional plain-film routines, series, or examinations are used to survey the temporal bone (Table 6.3) and to examine the paranasal sinuses (Table 6.4). A variety of radiographic signs can be recognized from alterations in the interfaces and contrasts among bone, soft tissue, and air in accordance with the 16 shades of gray recognized by the human eye. The radiographic signs of paranasal sinus disease are listed in Table 6.5.
Familiarity with conventional radiographic signs and a protocol for careful examination of radiographs based on proper clinical information leads to effective interpretation. The conventional temporal bone examination rarely is used today. Computed tomographic examinations are readily available and provide superior temporal bone information.
Acquisition of conventional x-ray films of the sinuses is the most common radiographic examination performed in the practice of otolaryngology. More than half of all conventional radiographic studies requested are of the paranasal sinuses, and these may be augmented by other radiographs, such as those obtained in the right and left orbital oblique projections and those of the nasal bones and zygomatic arches. Selective views, such as a lateral radiograph of the nasopharynx for measuring the adenoids or of the larynx and trachea for detecting a foreign body, are also commonly requested. These regional screening examinations and selective views should be reviewed to assure technical quality before the patient leaves the radiology department.
Some of the conventional imaging modalities listed in Table 6.2 rarely are requested because newer forms of imaging have more to offer. However, panoramic tomography of the teeth, mandible, and maxilla and a barium swallow examination of the hypopharynx and esophagus frequently are requested. Modified barium swallow is an actively monitored examination in which the radiologist’s attention must be focused on clear identification of the clinical problem to be solved.
High-technology Imaging
High-technology imaging (Table 6.6) includes CT, MRI, angiography, diagnostic ultrasonography, radionuclide scanning, and positron emission tomography. High-technology imaging supplements the screening information obtained with conventional radiographic examinations. The expansion of information occurs structurally. For example, CT expands the gray scale of conventional radiography into the thousands. Expansion also occurs physiologically, as in low-resolution radionuclide scans. Magnetic resonance imaging and angiography provide anatomic and physiologic information. High-technology imaging allows intervention such as ultrasound-guided fine-needle aspiration (FNA) biopsy or intra-arterial embolization.
HEAD AND NECK RADIOLOGY BY REGION
Temporal Bone
The temporal bone is the most complex anatomic structure in the body, and pathologic changes may induce only modest radiologic signs. The proximity of the temporal bone to important bony structures, such as the base, vault, and foramina of the skull, leads to summation and superimposition, which cause masking effects of fine details on conventional images. Radiologic investigation of the temporal bone complex and related structures may be inadequate unless high-technology imaging modalities are used (Table 6.6). Before sophisticated imaging modalities were developed, standard projections (Table 6.3), pneumoencephalography, contrast cisternography, and, beginning in the 1950s, complex motion or pluridirectional polytomography were widely used for the evaluation of temporal bone disease.
Conventional temporal bone projections still are used in many parts of the world where CT and MRI are not readily available. Conventional modalities depict the key attic, aditus, and antral region, mastoid pneumatization, and petrosa. Some pathologic conditions, such as chronic suppurative otitis media with or without cholesteatoma, acute coalescent mastoiditis, and large tumors of the internal auditory canal or the cerebellopontine angle such as acoustic neuroma, meningioma, and subarachnoid cyst, can be identified primarily from evidence of bone destruction. Pneumoencephalography and cisternography were used mainly to depict acoustic neuroma by using contrast medium to accentuate filling defects. Complex motion tomography supplemented these interventional procedures. These techniques allowed multiplanar assessment of bone destruction. However, acoustic neuroma was identified from bone destruction in only 60% of instances. A tumor had to reach a large size to produce enough bone destruction to be radiologically recognized. Complex motion tomography allowed assessment of congenital malformations, otosclerosis, and temporal bone fractures.
Computed tomography can be performed with high-resolution bone algorithms, and a series of contiguous or overlapping CT sections can be processed to generate excellent images. Spatial resolution in MRI has increased such that information offered about the cerebellopontine angle, internal auditory canal, cochlea, and vestibule exceeds that obtained with CT. A contrast-enhanced study is of value in many situations. Magnetic resonance imaging and CT are not mutually exclusive examinations and are frequently combined to fully assess lesions of the base of the skull. Angiography is reserved for the detection of arterial stenotic lesions and arteriovenous fistulae, for the evaluation of tumor vascularity, and for preoperative embolization.
Computed tomography and MRI have become the radiologic methods of choice for most disorders of the ear and temporal bone. These high-technology studies also can be used to evaluate the neighboring skull base and related structures of the middle and posterior cranial fossae. Magnetic resonance imaging and CT are complementary. Computed tomography is good for imaging diseases that affect cortical bone, air spaces, and some soft tissue lesions of the temporal bone. Magnetic resonance imaging is excellent for evaluating soft tissues, cerebrospinal fluid, and blood vessels. However, the thin cortical bone and air-containing spaces of the middle ear and mastoid cause signal voids, which make it difficult to differentiate air from bony septations. Therefore, high-resolution CT is still the procedure of choice for bone assessment. Computed tomography is excellent for evaluating congenital anomalies, including cochlear and labyrinthine anomalies, with a special emphasis on Mondini dysplasia, congenital microtia, and other middle ear abnormalities. The degree of otomastoid pneumatization, the status of the ossicular chain, the course and position of the facial nerve, the positions of the internal carotid canal, sigmoid sinus, and jugular bulb, the thickness of the atretic plate, and the dimensions of the middle ear all are important in assessing operability. In most cases, intravenous administration of contrast material is not needed unless the study is performed to evaluate the central nervous system or vascular structures. The main disadvantages of CT are that it cannot be used to image the internal auditory canal and that it has limited value in imaging in multiple planes because many patients cannot be positioned for direct coronal or sagittal images.
In the diagnosis of erosive disease of bone, such as chronic otitis media with cholesteatoma, CT with bone window settings provides the best information. The locations, extent, and nature of the complications of cholesteatoma are ideally evaluated with CT. Acute coalescent mastoiditis, which usually occurs as masked mastoiditis, can be identified by means of thin-section axial and coronal CT examinations. Intracranial complications, such as sigmoid sinus thrombosis, extradural and subdural abscesses, and brain abscess, are best detected with contrast-enhanced CT with both bone and soft-tissue window settings. Even with MRI, dural sinus thrombosis is a difficult imaging diagnosis. Magnetic resonance angiography and MR venography are state-of-the-art examinations for evaluation of this dangerous clinical condition. Computed tomography allows the relative identification and differentiation of soft-tissue abnormalities associated with chronic ear disease, such as granulation tissue, mucopurulent effusions, cholesteatoma, and cholesterol granuloma. High-resolution CT is an excellent and sensitive means of imaging temporal bone fracture lines. Magnetic resonance imaging is recommended if the presence of coexisting intracranial abnormalities is suspected. Hemotympanum, ossicular chain disruption, and injury to the facial nerve canal are common radiologic findings. Posttraumatic cerebrospinal fluid leak is still best depicted on intrathecally enhanced CT scans.
Computed tomography with intravenous administration of contrast medium has been used routinely in screening for masses in the cerebellopontine angle. Acoustic neuroma is the most frequent tumors in this location. Computed tomography can accurately depict lesions larger than 1 cm in diameter, but it is generally unreliable for smaller tumors. Since the introduction of paramagnetic contrast agents, the usefulness of MRI has greatly improved. When a retrocochlear lesion is suspected on clinical grounds, MRI with gadolinium enhancement is the imaging modality of choice (1). It is superior to CT for evaluating acoustic neuroma because with contrast medium it can depict tumors smaller than 0.8 cm in diameter. Magnetic resonance imaging with gadolinium–diethylenetriamine pentaacetic acid contrast enhancement can help in the assessment of primary soft-tissue abnormalities in the temporal area, such as facial nerve lesions, labyrinthine schwannoma, and vestibular neuronitis. The superiority of MRI also is recognized in the evaluation of other lesions that affect the cranial nerves and cerebrospinal axis (2).
Other diagnostic modalities for vascular imaging of the temporal bone area include superselective angiography and digital subtraction angiography for screening. Angiography also can be used to direct interventional procedures. These modalities are useful in evaluating primary vascular tumors, such as glomus jugulare tumor and its extension beyond the jugular foramen into the skull base and temporal bone. These studies also are important in recognizing anomalies that can lead to surgical disasters. An aberrant middle ear internal carotid artery or high jugular bulb can be identified by means of primary vascular screening studies, by means of reconstructive or direct CT, and by means of MR angiography.
Combination radionuclide bone and gallium scans are essential in diagnosing malignant external otitis (osteomyelitis of the temporal bone) and in assessing stage and response to treatment. These scans provide physiologic information beyond the morphologic findings on CT scans.
Paranasal Sinuses
The paranasal sinuses are air-filled cavities surrounded by bone. They are inaccessible to direct clinical examination unless telescopic intervention is used. Cooperation between clinician and imager is essential for effective treatment decisions and follow-up evaluation, especially for endoscopic techniques, in which the osteomeatal complex must be thoroughly studied with CT.
Conventional plain radiography once was the imaging modality of choice in the evaluation of the paranasal sinuses. The clinical and radiographic emphasis was on disease in the maxillary and frontal sinuses. Superimposition of structures precluded accurate evaluation of the anatomic relations of the ethmoid sinuses and osteomeatal complexes. With the recognition of the importance of the osteomeatal complex in sinus disease and a change in therapeutic approach, CT has replaced conventional radiography as the primary diagnostic modality.
The target structures and views for conventional plain radiographic examination of the paranasal sinuses are summarized in Table 6.4. A complete series of sinus radiographs usually includes Waters, Caldwell, lateral, and submentovertex views; right and left oblique orbital views are obtained if the orbital apex or optic canal approximates the area of concern. Additional views may be needed to study the nasal bones or zygomatic arches. Panoramic tomography or focused dental radiographs can be used to evaluate apical dental disease that affects the maxillary sinuses. Conventional radiographs can be accurate in showing air-fluid levels, but the degree of chronic inflammatory disease present is consistently and substantially underestimated (3). Unlike conventional radiography, CT clearly depicts the fine bony anatomy of the osteomeatal complexes.
Before the introduction of CT, complex motion tomography provided critical anatomic information, especially when bone destruction or bone displacement was in question. Contiguous sections can be as close as 1 mm apart in coronal or lateral projections to detail regions of interest after a preliminary screening examination has been performed with 3- to 5-mm contiguous sections, as in assessing a blow-out fracture of the orbital floor. This technique is used if CT is not available. Although it allows clear morphologic visualization of bony abnormalities, the conventional technique has a limited gray scale.
Computed tomography has been advocated for improved diagnosis and is considered the best method for evaluating the paranasal sinuses. Imaging in the coronal plane is recommended because it optimally displays the osteomeatal units, including the relation of the brain to the ethmoid roof and the relation of the orbit to the paranasal sinuses. Coronal images closely correlate with the surgical approach used in endoscopic sinus surgery. Axial images are recommended in addition to coronal images when a patient has severe disease in the frontal, sphenoid, or posterior ethmoid sinuses, especially if surgery in these regions is being considered. The initial CT scan should be obtained after an adequate course of medical therapy to eliminate reversible mucosal inflammation (4). Modern spiral CT scanners rapidly acquire thin axial images. Use of reformatted coronal images allows examination of pediatric patients and of patients who find it difficult to maintain the position needed for coronal imaging.
Inflammatory conditions, trauma, mucocele, and tumor are critical conditions in which the use of CT is mandatory. Understanding orbital-sinus and cranial-sinus relations is essential in assessing the pretreatment stage of carcinoma of the nasal cavities or paranasal sinuses, response to radiation therapy, or postoperative tumor recurrence. Axial examinations are essential for observing breaching of the posterior wall of the maxillary sinus. Although CT can be performed without contrast enhancement, enhancement can help differentiate obstructive secretions and a mass (MRI is most useful for this purpose). Tumor extension beyond the sinuses into the orbit, brain, or retromaxillary region is best seen with contrast enhancement.
Contrast enhancement techniques are important in evaluating vascular lesions. Angiofibroma of the nasopharynx necessitates use of this technique to identify the epicenter and extension of angiofibroma from the region of the sphenopalatine foramen, medially into the nasopharynx or laterally into the infratemporal fossa and anteriorly into the maxillary sinus or superiorly into the middle cranial fossa. Cerebrospinal fluid leak and meningocele are two neurogenic disorders that can necessitate intrathecal CT enhancement. Computer-generated three-dimensional CT images are being used increasingly in evaluating complex facial fractures and severe craniofacial anomalies. Computed tomography and MRI can be used intraoperatively during image-guided sinus surgery. Conventional radiographs, complex motion tomography, and CT have common radiographic signs (Table 6.5) (5).
Magnetic resonance imaging is most useful in the evaluation of regional and intracranial complications of inflammatory sinus disease, in the assessment of benign versus malignant sinus opacification. and in the evaluation of the extent of neoplastic processes (6,7). Compared with CT, MRI provides better visualization of soft tissue, but it does not optimally display the cortical air-bone interface. Therefore, CT is still a more reliable operative “road map” for a surgeon performing an endoscopic procedure on the sinuses. Staging systems are used to evaluate CT scans of the sinuses (8).
The introduction of MRI has influenced the diagnosis, management, and follow-up evaluation of tumors of the nasal cavities, paranasal sinuses, and nasopharynx. Magnetic resonance imaging is better than CT in the assessment or characterization of soft-tissue mass lesions. Magnetic resonance imaging depicts vascular structures without the use of intravenous contrast agents. Inflammatory lesions have high signal intensity on T2-weighted images, and tumor masses tend to have low or intermediate T2-weighted signal intensity. If MRI shows a noninflammatory mass, biopsy should be strongly considered.
Other diagnostic modalities used to study the paranasal sinuses have lesser but selective importance. Ultrasonography of the maxillary and frontal sinuses has limited usefulness but can help identify fluid when conventional radiographs show opacification. Radionuclide bone scans with technetium 99m methylene diphosphonate depict osteoblastic activity in several situations, particularly identification of focal manifestations of systemic disease, such as Paget disease. Detection of these nonspecific findings is highly sensitive, and the bone scan technique shows physiologic changes before morphologic studies with conventional radiographs and CT scans do. Gallium citrate scans, if used in combination with bone scanning, allow diagnosis of osteomyelitis and follow-up assessment of the effectiveness of treatment. Treatment is considered successful when results of a gallium scan return to normal. This modality specifically images the infective focus, unlike bone scanning, which images the osteoblastic response around the infective focus (9).
Soft Tissues of the Neck
Head and neck masses have traditionally been classified as benign or malignant, primary or metastatic, and congenital or inflammatory. This classification is further extended to cover age groups, such as children and adults, and location, such as midline and lateral, anterior triangle and posterior triangle. Although the most important preliminary step in evaluating neck masses is a careful physical examination of the neck and all mucosal surfaces, this often only establishes a working diagnosis or defines an outstanding clinical problem (Table 6.1). The uncertainty associated with clinical diagnosis is determined by limitations in differentiating solid and cystic lesions and in determining anatomic associations. Radiologic imaging is essential in responding to these clinical problems.
Conventional plain radiographs are not usually helpful in differentiating neck masses, except for recognizing infrequent signs such as calcification. Ultrasonography is a safe, relatively inexpensive, and readily available investigative method. Ultrasonography is categorized as high-resolution real-time imaging or as Doppler-based technique used to assess flow characteristics of the major blood vessels of the neck. Palpable lumps in the neck can be assessed with ultrasonography, which can be used for specific assessment of growth (size of lesion over time); location; relation of lesion to the adjacent structures, especially blood vessels; character of the lesion (solid, cystic, complex); and the number and size of affected lymph nodes in the region. Applications of ultrasonography include assessment of thyroglossal duct abnormalities, branchial cleft cysts, cystic hygroma, major salivary gland tumors, inflammatory processes that progress to abscess formation, and carotid body tumors, lymph node staging, and imaging-guided FNA.
Ultrasound techniques combined with FNA and cytologic evaluation have particular importance in the assessment of the soft tissue of the neck (10). The superiority of ultrasonography is only in its nonradiographic guidance capability including FNA biopsy of cervical lymph nodes or other masses. Ultrasonography also can be useful for percutaneous drainage of cervical abscesses. Fine needle aspiration biopsy guided by computed tomography can be used to diagnose poorly accessible or deep-seated lesions of the head and neck (11).
Computed tomography and MRI are widely used for primary staging of tumors and nodes. However, accuracy in assessing lymph nodes depends on the radiologic criteria used. Various criteria for the size of lymph nodes have been used for diagnosing metastatic lymphadenopathy, but the most reliable imaging finding is the presence of nodal necrosis. Areas of central nodal necrosis larger than 3 mm are routinely identified on contrast-enhanced CT scans. The usefulness of MRI and CT in the detection of lymph node metastasis from squamous cell carcinoma of the head and neck was assessed by Curtin et al. (12). Computed tomography performed slightly better than MRI in the detection of lymph node metastasis. A high negative predictive value was achieved only when a low size criterion (5 to 10 mm) was used and was therefore associated with a relatively low positive predictive value. For example, with criteria of 1-cm size or an internal abnormality to indicate a positive node, CT had a negative predictive value of 84% and a positive predictive value of 50%. Magnetic esonance imaging had a negative predictive value of 79% and a positive predictive value of 52%. Ultrasonography is hampered by similar morphologic criteria, and only ultrasound-guided FNA biopsy can offer additional cytologic criteria that may be more reliable (13). Positron emission tomography with 18-fluorodeoxyglucose in the detection of cervical lymph node metastasis has been shown in several studies (14,15) to have a higher sensitivity and specificity than CT, MRI, or ultrasound-guided FNA biopsy.
Compared with MRI, CT is the best method for studying capsular penetration and extracapsular nodal extension. With contrast CT, extracapsular extension is identified with an enhancing nodal rim, usually with infiltration of the adjacent fat planes. Carotid artery invasion is an important prognostic indicator, and invasion of the adventitia is as important as greater degrees of arterial invasion. Detection of microscopic adventitial infiltration is beyond the scope of current imaging, but circumferential involvement of the artery by tumor is strong evidence that the artery is invaded.
Both CT and MRI are useful to evaluate for tumor recurrence and posttreatment changes in the care of patients with malignant tumors of the head and neck. Magnetic resonance imaging has advantages in differentiation between tumor and scar tissue, but edema after radiation therapy can make differentiation difficult. Radiation-induced changes lead to false-positive diagnoses for approximately 50% of patients (16). Criteria for recurrent or residual tumor include an infiltrative mass with high signal intensity on T2-weighted images and enhancement after administration of gadolinium on T1-weighted images. Computed tomographic criteria for recurrent or residual tumor include the combination of a circumscribed, infiltrative mass with contrast enhancement on CT scans. Positron emission tomography compares favorably with CT and MRI in the detection of recurrent or residual cancer (17).
Other methods for detecting tumor recurrence include thallium 201 single photon emission computed tomography (SPECT) and gallium 67 citrate whole-body scintigraphy. Mukherji et al. (1
found thallium 201 SPECT to be superior to CT for differentiating recurrent tumor from posttreatment changes. Murata et al. (19) found gallium 67 citrate whole-body scintigraphy especially useful for evaluation for recurrence and distant metastasis of squamous cell carcinoma of the head and neck.
Computed tomography is useful for detecting clinically occult primary cancer of the head and neck that causes cervical metastasis. Magnetic resonance imaging can better depict submucosal lesions in the areas of the tonsils, anterior floor of mouth, and base of the tongue. Magnetic resonance imaging should be performed first for a mass in the retropharyngeal or parapharyngeal space. In these locations, tumors of neural, vascular, or salivary origin should be suspected. Magnetic resonance imaging also should be performed if posterior extension into the airway, esophagus, or posterior deep muscles is suspected. With few exceptions, CT is better than MRI in depicting thyroglossal duct cysts, branchial cleft cysts, and cystic hygroma and in differentiating infection, inflammatory processes such as cellulitis, edema, and abscess.
Digital subtraction angiography and conventional superselective angiography are useful in the diagnosis of hemangioma, arteriovenous malformation, and paraganglioma. Intra-arterial embolization has important therapeutic applications as definitive treatment or before surgery.
Despite being hypovascular, nerve sheath tumors can become greatly enhanced on CT scans, apparently because of extravascular leakage of contrast material into the tumor bed. Nerve sheath tumors have intermediate signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Paraganglioma becomes intensely enhanced after intravenous administration of contrast material during CT and MRI studies. The characteristic salt-and-pepper appearance on MR images reflects the signal voids of many tumor vessels. However, signal voids do not occur with all paragangliomas, especially in those less than 2 cm in diameter.
Larynx
Radiologic imaging modalities are important in screening and in defining the deep dimension of a malignant tumor of the larynx. Although the larynx is readily accessible to direct visualization, including telescopic assessment, and biopsy, submucosal extension is not subject to direct visualization. If possible, imaging studies should be performed before biopsy to avoid confusion of tumor and local trauma. Vocal cord mobility defects, whether due to direct infiltration by tumor or involvement of the recurrent laryngeal nerve, can be assessed with diagnostic imaging. There also are difficult-to-evaluate regions, such as the Morgagni ventricle or the subglottis. Imaging is critical in the evaluation of carcinoma of the larynx in all three major regions (supraglottis, glottis, subglottis) and in evaluation of extralaryngeal extension of malignant growth to the hypopharynx or the laryngeal cartilage. Virtual endoscopy of the airway is possible, but it has limited usefulness in differentiating mucosal surfaces that are touching (20).
Conventional plain radiographs (lateral and anteroposterior projections and selective high-kilovoltage filtration techniques) of the larynx provide preliminary or definitive information about foreign bodies, trauma, and other types of acute and chronic airway obstruction. These radiographs can show soft-tissue swelling, alteration of the cartilaginous framework if sufficiently calcified, and the position of the air column. The variability of calcification of the laryngeal cartilage can pose a diagnostic problem in the detection of foreign bodies. For example, the superior margin of the cricoid cartilage calcifies long before the signet portion does. The result is linear calcification on plain radiographs that often is mistaken for a foreign body.
Xeroradiography, because it provides edge enhancement, can clarify intrinsic soft-tissue detail such as calcification, delineate masses and stenosis, sometimes depict cartilage abnormalities such as fractures and erosions, and help identify foreign bodies by type and location. Unfortunately, this technique carries a radiation exposure three to five times that of conventional radiography. The usefulness of xeroradiography in imaging of the soft tissue and cartilage has been superseded by that of CT and MRI.
Conventional coronal tomography allows visualization of frontal view anatomy without a superimposed spine. Thus it allows satisfactory analysis of the vertical extent of laryngeal tumor and subglottic or tracheal stenosis or stricture. This technique has been replaced by CT and MRI because of its limited gray scale for soft-tissue differentiation, but the airway image, especially with the added sagittal projection, is excellent. Tomography has several limitations, such as poor definition of the anterior commissure. Depiction of cartilage invasion is unreliable, except when there is extensive involvement of well-calcified cartilage.
Ultrasonography has limitations because the laryngeal cartilage reflects most of the sound, limiting ultrasonic access. Nuclear medicine imaging seldom is useful for laryngeal imaging. Increased uptake has been mentioned as a cause of inflammatory arthropathy and relapsing polychondritis. Erythrocyte-tagged imaging can provide enough information for diagnosis of the rare laryngeal cavernous hemangioma. Arteriography is seldom used except for evaluation of a suspected vascular lesion such as paraganglioma.
Technologic advances in CT and MRI have greatly improved the ability to image the larynx. Spiral CT and fast MRI techniques allow rapid acquisition, which decrease degradation motion artifacts from breathing, swallowing, and carotid artery pulsation. Both CT and MRI allow evaluation of the extent of laryngeal tumors, especially for tumor size staging of carcinoma. Such determinations can influence the extent of laryngectomy (partial versus total).
Spiral CT scanners acquire the complete data set through the larynx in less than 10 seconds, allowing the patient to stay motionless. Images can then be reconstructed to give overlapping sections, and coronal, sagittal, and even three-dimensional images can be generated from the same data set. Examination of the larynx during various inspiratory-expiratory cycles has been used to optimize visualization of a particular region or the margin of a tumor. With operator-directed imaging, CT provides accurate images of the location, size, and extent of the tumor. Computed tomography can depict cartilage invasion and deep soft-tissue spread into the paraglottic space, preepiglottic space, and pyriform sinuses. This T-stage imaging contributes to a more rational diagnosis and more effective planning of surgical and radiation treatment.
The larynx has been difficult to image well with MRI because of motion artifact. Fast spin echo imaging has made a marked difference in the ability of MRI to image the larynx by reducing these motion artifacts. The use of gadolinium is controversial. Some experts believe gadolinium enhancement provides important information regarding the interface of tumor with muscle. However, fast spin echo imaging can generate similar information, obviating administration of gadolinium. Magnetic resonance imaging is better than CT at separating soft tissues. Another advantage of MRI over CT is acquisition of high-resolution images in multiple planes. Sagittal images show the epiglottis, vallecula, and base of the tongue well. Coronal views are ideal for evaluating the margins of the vocal cords, the paraglottic space, and the vertical extent of tumor. Axial images allow assessment of cartilaginous erosion.
The appearance of the cartilage on CT scans and MR images varies with the degree of ossification, which is not uniform and frequently is asymmetric. Invasion of cartilage has implications in staging and outcome of carcinoma of the larynx. With CT, tumors often have the same attenuation as nonossified cartilage (soft-tissue density), so minimal cartilage involvement can be difficult to assess. Gross cartilage destruction with tumor on the opposite side of the cartilage from the primary lesion is the only truly reliable CT sign of cartilaginous invasion. This finding is similar in MRI, but there is evidence to suggest that lesser degrees of tumor involvement can be detected with MRI because of variability in the appearance of both normal and abnormal cartilage afforded by MRI sequences. Nodal staging can be dramatically improved with sensitive broadening of the primary CT examination. Thin-section CT added to the primary tumor assessment can show enlarged or metastatic nodes in the neck and specific anatomic and pathologic features.
Salivary Glands
The diagnosis of salivary gland disorders is established from the findings of the history, physical examination, and FNA biopsy rather than from radiographic studies. However, imaging studies often are needed to assess focal, multifocal, diffuse, and bilateral disorders. New diagnoses, such as stones, can be readily established. Imaging studies can be used to confirm a clinical suspicion, such as human immunodeficiency virus–positive status from parotid gland enlargement with lymphoepithelial cysts.
Conventional radiographic examination that includes occlusal radiography and panoramic tomography can depict radiopaque duct calculi. Most submandibular duct and gland stones are radiopaque. Parotid duct stones occur infrequently and usually are radiolucent. Conventional sialography once was the standard method of assessment of the morphologic features of the salivary duct and glands. It is not commonly performed now, although it provides important information about nonopaque stones, ductal stenosis, sialectasis, and sialosis.
In imaging of the major salivary glands, ultrasonography has reasonable accuracy in differentiating intracapsular and extracapsular lesions. It also images solid and cystic lesions. Mixed solid-cystic lesions, such as Warthin tumor, are considered solid. Ultrasonography is limited in depicting deep-lobe parotid lesions because of the attenuation and reflection of sound by the mandible.
Radionuclide scanning is useful in some pathophysiologic salivary gland assessments because 99mTc-pertechnetate is incorporated into the salivary glands and excreted in the saliva. Increased focal uptake is characteristic of functioning tumors, such as Warthin tumor of the parotid gland and the rare oncocytoma. Diffuse, increased radionuclide uptake often indicates ductal obstruction leading to intra-acinous salivary retention in the presence of long-standing inflammation. Positron emission tomography does not reliably differentiate benign from malignant tumors, limiting its clinical usefulness. It also is expensive and takes a long time to perform.
Computed tomography and MRI are excellent methods for imaging and evaluating salivary gland disease, specifically focal, multifocal, and diffuse masses. The two modalities have equivalent diagnostic potential for imaging solid and cystic lesions. Both provide important information about the location (intraglandular or extraglandular), size, and extension of tumor to surrounding superficial and deep structures.
Bone invasion is viewed differently with MRI and CT, and marrow invasion of the mandible is better identified with MRI. The excellent soft-tissue depiction with MRI allows identification of the intraparotid portion of the facial nerve and its relation to infiltrative masses, which allows planning of surgical management. Magnetic resonance imaging also is superior to CT in evaluating the muscle-tumor interface. Another advantage of MRI over CT is the absence of exposure to radiation or the intravenous administration of iodine-containing contrast medium.
The choice of which imaging study to perform in investigating salivary gland disease is influenced by clinical presentation, user preference, and familiarity with a specific modality. Computed tomography is the choice in cases of inflammatory disease of the salivary glands, because MRI does not depict ductal dilatation or salivary calcification. However, if the clinical finding is a mass, the initial imaging evaluation, if any, usually is MRI. Computed tomography is an acceptable alternative, and ultrasonography can be used as a complementary study.
Thyroid Gland
The superficial position of the thyroid gland in the neck enables easy access for clinical examination and FNA. A wide range of high-technology imaging modalities are available for the diagnosis and management of thyroid disease. These may generate structural information, as do ultrasonography, CT, and MRI, or show tissue function, as does radionuclide scanning (21).
Conventional radiography is not a primary study. It is limited to screening for airway or esophageal displacement or invasion and to identification of calcification in the thyroid gland. Radionuclide scanning often is chosen for imaging malignant lesions of the thyroid. Three radioactive pharmaceuticals are commonly used in clinical practice. Sodium 99mTc-pertechnetate is trapped by the thyroid gland but is not organified. Radioiodine isotopes (iodine 123 and iodine 131) are trapped and organified by the parenchyma of the thyroid gland. Pertechnetate is the most commonly used radioisotope, at least for the initial evaluation. It is less expensive than radioisotopic iodine, is readily available, and approximates iodine trapping through the metabolism of the pertechnetate anion, which is not incorporated into hormonogenesis. Because 123I radioiodine scanning is unique in providing anatomic images and images of the functional activity of the thyroid gland or ectopic thyroid tissue, it is useful in a variety of clinical situations. These include investigation of a palpable thyroid nodule or a mass in the midline of the neck, base of the tongue, or mediastinum. Radiopharmaceuticals can be used in the management of cancer and Graves disease and in screening for thyroid metastasis and postsurgical recurrent tumor.
Radionuclide scanning has several limitations. The anatomic imaging resolution is only 1.0 cm, which restricts detail and definition. Scans may not give an adequate image of thyroid tissue if the patient is taking oral thyroid hormone supplements. A hot nodule on a sodium 99mTc-pertechnetate scan may or may not indicate a functioning tumor. In these situations, a radioactive iodine (123I) scan shows physiologic hormonogenesis and a functioning or nonfunctioning mass.
High-resolution ultrasonography is the first-line structural investigative modality in the diagnosis of many thyroid disorders, especially nodular disease. It is safe, inexpensive, simple, quick, and reproducible. Marked improvement in image quality occurred with the introduction of small-parts ultrasonography. Ultrasonography has an accuracy greater than 90% in differentiating cystic and solid thyroid nodules. Mixed solid-cystic nodules should be managed as solid masses. Ultrasonography can depict questionable or difficult-to-palpate lesions and can be used to direct FNA. It also can show whether a palpable nodule is part of a focal, multifocal (multinodular goiter), or diffuse process. Ultrasonography can be used to follow the size of a nodule after suppression therapy or cyst aspiration. The major limitation of ultrasonography is the inability to differentiate malignant from benign lesions on the basis of tissue characteristics. This is appropriately left to a pathologist. Retrosternal thyroid cannot be evaluated because of the bony interference with sound.
Computed tomography and MRI have similar roles in the evaluation of thyroid disorders and are essentially second-line imaging modalities for this region of the neck. Both techniques provide useful information about the size, shape, and anatomic structure of thyroid nodules. They can help determine whether a mass is solitary or part of a multinodular lesion. Computed tomography and MRI can be used to evaluate mediastinal, substernal, or retrosternal extension of thyroid masses and regional lymph node involvement or local recurrence. Computed tomography shows calcification better than does MRI, but MRI is superior in providing soft-tissue detail, especially for the muscle-tumor interface. Magnetic resonance imaging can be performed without intravenous contrast medium and does not expose the patient to radiation. If iodinated contrast material is used during CT, the iodine interferes for months with thyroid function tests and uptake of iodine 131 used postoperatively to manage well-differentiated malignant disease of the thyroid.
Parathyroid Glands
Parathyroid imaging is controversial in terms of the indications for imaging and the imaging agent used. Preoperative imaging localization studies are not performed at many centers if the patient has not undergone surgical intervention (22). There are many options for imaging the parathyroid glands, including ultrasonography, CT, MRI, angiography and many nuclear medicine studies. Ultrasonography is the least invasive. Because of their small size, normal parathyroid glands are not usually detected with ultrasonography. However, parathyroid lesions larger than 0.5 cm in diameter usually can be identified in a careful study. In cases of hyperparathyroidism due to parathyroid adenoma, accurate localization is reported at a rate between 69% and 88% (23). Our success rate is greater than 90%. For perithyroidal parathyroid adenoma, ultrasonography is an excellent imaging choice, but acoustic impedance prevents adequate imaging behind air-filled structures (trachea) or bone (mediastinum).
No nuclear medicine agent is exclusively taken up by normal or adenomatous parathyroid glands. The agents that are taken up by both parathyroid adenoma and thyroid tissue (thallium 201 and 99mTc-sestamibi) are subtracted from agents that are taken up by thyroid tissue only (99mTc-pertechnetate and iodine 123). However, 99mTc-sestamibi imaging is commonly performed without subtraction because this agent washes out of the thyroid gland rapidly but is retained by parathyroid tissue and thyroid adenoma. Although these scans have limited value in detecting four-gland hyperplasia but have some usefulness in detecting asymmetric hyperplasia, diagnostic accuracy is reported at between 50% and 95% in cases of parathyroid adenoma (24).
Digital subtraction angiography relies on a single morphologic characteristic—the regional blood supply to a hypervascular gland. After the intra-arterial or intravenous injection of contrast medium, a computerized image-subtraction program eliminates unnecessary background information. This technique allows identification and localization of parathyroid adenoma in 60% to 70% of patients. This modality rarely is needed because sestamibi scans and ultrasonography are so successful. It may be useful in revision cases of hyperparathyroidism.
Normal parathyroid glands seldom are depicted on CT scans. Markedly enlarged glands can be detected with CT with a sensitivity of 50% to 88%. In the neck, an axial CT image is no more informative than ultrasonography. Computed tomography is limited by technical factors, such as inadequate resolution of lesions less than 1.0 cm in diameter, or by limitations in interpretation, such as mistaking a tortuous vessel, thyroid mass, or lymph node for an enlarged parathyroid gland. Use of iodinated contrast agents for differentiating blood vessels from adenoma or lymphadenopathy prevents subsequent imaging with iodine-based nuclear medicine studies for approximately 6 weeks.
Normal parathyroid glands usually are not identified with MRI. Magnetic resonance imaging with gadolinium enhancement can be useful for evaluating hyperparathyroidism refractory to the first surgical attempt at correction. T2-weighted MRI and gadolinium fat-suppressed T1-weighted scans show parathyroid adenoma as an area of high signal intensity against a dark, soft-tissue background. A review of the parathyroid literature up to 1993 showed that MRI had the highest sensitivity for the detection of adenoma (74%) followed by nuclear medicine studies (72%), CT (65%), and ultrasonography (63%) (25). At reoperation for previously unidentified adenoma, MRI had the highest (66%). Ultrasonography had a sensitivity of 60%; CT, 48%; and nuclear medicine studies, 45%.
PRINCIPLES OF IMAGING
A diagnostic imaging study is the most common consultation that otolaryngologists and head and neck surgeons request for a patient. Because many of the supporting and deep structures of the head and neck are beyond direct topographic evaluation, even with fiberoptic endoscopes and telescopes, further anatomic and physiologic information must be obtained for the temporal bone, skull base, paranasal sinuses, soft tissues of the neck, larynx, and other structures by means of diagnostic imaging. This does not mean that every patient needs an imaging study. Some patients need only a sensitive and thoughtful examiner. Other patients need conventional plain radiographs to complete the assessment. Others, however, need high-technology imaging for an accurate diagnosis and optimal treatment planning.
The influence of professional development and the team approach to diagnostic imaging is growing. Neuroradiologists have long gravitated to the field of head and neck radiology, and interventional specialization has extended far beyond this discipline. Angiography provides a model in which diagnostic imaging has led to interventional imaging and the training of dedicated specialists. A radiologic examination must be treated as a consultation to derive optimal information for formulating the best working diagnosis. This is the essence of effective diagnostic imaging, and it is achieved through effective communication between otolaryngologist and radiologist and through mutual analysis of the clinical problem and radiologic evidence. Consultation involves a well-prepared requisition, a telephone call, or a copy of the referring letter or consultation report.
The consulting team members must establish the optimal working diagnosis. Treatments may vary, but there is only one diagnosis, and it must be qualitatively and quantitatively effective (Table 6.1). Qualitative aspects of diagnosis usually lie within the field of the pathologist. However, the diagnostic image can provide qualitative answers to questions about whether disease exists, whether there is a specific disease process, such as a fracture, or whether a group disorder, such as bone destruction or cyst formation, can be identified. Diagnostic images also contain answers to quantitative questions about the extent of disease in all three dimensions, about disease that crosses regional boundaries that affect management, and about whether the disease is a local manifestation, metastatic condition, or systemic disorder. Imaging can help to inform the clinician about the presence and nature of a specific anatomic or physiologic derangement.
The foregoing are the types of issues clinicians formulate for imagers and for which solutions can be provided. However, routine questions receive only routine responses. Fifty percent of diagnostic information can be lost from computed tomography (CT) and magnetic resonance imaging (MRI) if the studies are not tailored to provide specific answers to pathologic anatomic and physiologic questions.
IMAGING MODALITIES
Conventional Imaging
The modalities representative of conventional imaging are listed in Table 6.2. Regional plain-film routines, series, or examinations are used to survey the temporal bone (Table 6.3) and to examine the paranasal sinuses (Table 6.4). A variety of radiographic signs can be recognized from alterations in the interfaces and contrasts among bone, soft tissue, and air in accordance with the 16 shades of gray recognized by the human eye. The radiographic signs of paranasal sinus disease are listed in Table 6.5.
Familiarity with conventional radiographic signs and a protocol for careful examination of radiographs based on proper clinical information leads to effective interpretation. The conventional temporal bone examination rarely is used today. Computed tomographic examinations are readily available and provide superior temporal bone information.
Acquisition of conventional x-ray films of the sinuses is the most common radiographic examination performed in the practice of otolaryngology. More than half of all conventional radiographic studies requested are of the paranasal sinuses, and these may be augmented by other radiographs, such as those obtained in the right and left orbital oblique projections and those of the nasal bones and zygomatic arches. Selective views, such as a lateral radiograph of the nasopharynx for measuring the adenoids or of the larynx and trachea for detecting a foreign body, are also commonly requested. These regional screening examinations and selective views should be reviewed to assure technical quality before the patient leaves the radiology department.
Some of the conventional imaging modalities listed in Table 6.2 rarely are requested because newer forms of imaging have more to offer. However, panoramic tomography of the teeth, mandible, and maxilla and a barium swallow examination of the hypopharynx and esophagus frequently are requested. Modified barium swallow is an actively monitored examination in which the radiologist’s attention must be focused on clear identification of the clinical problem to be solved.
High-technology Imaging
High-technology imaging (Table 6.6) includes CT, MRI, angiography, diagnostic ultrasonography, radionuclide scanning, and positron emission tomography. High-technology imaging supplements the screening information obtained with conventional radiographic examinations. The expansion of information occurs structurally. For example, CT expands the gray scale of conventional radiography into the thousands. Expansion also occurs physiologically, as in low-resolution radionuclide scans. Magnetic resonance imaging and angiography provide anatomic and physiologic information. High-technology imaging allows intervention such as ultrasound-guided fine-needle aspiration (FNA) biopsy or intra-arterial embolization.
HEAD AND NECK RADIOLOGY BY REGION
Temporal Bone
The temporal bone is the most complex anatomic structure in the body, and pathologic changes may induce only modest radiologic signs. The proximity of the temporal bone to important bony structures, such as the base, vault, and foramina of the skull, leads to summation and superimposition, which cause masking effects of fine details on conventional images. Radiologic investigation of the temporal bone complex and related structures may be inadequate unless high-technology imaging modalities are used (Table 6.6). Before sophisticated imaging modalities were developed, standard projections (Table 6.3), pneumoencephalography, contrast cisternography, and, beginning in the 1950s, complex motion or pluridirectional polytomography were widely used for the evaluation of temporal bone disease.
Conventional temporal bone projections still are used in many parts of the world where CT and MRI are not readily available. Conventional modalities depict the key attic, aditus, and antral region, mastoid pneumatization, and petrosa. Some pathologic conditions, such as chronic suppurative otitis media with or without cholesteatoma, acute coalescent mastoiditis, and large tumors of the internal auditory canal or the cerebellopontine angle such as acoustic neuroma, meningioma, and subarachnoid cyst, can be identified primarily from evidence of bone destruction. Pneumoencephalography and cisternography were used mainly to depict acoustic neuroma by using contrast medium to accentuate filling defects. Complex motion tomography supplemented these interventional procedures. These techniques allowed multiplanar assessment of bone destruction. However, acoustic neuroma was identified from bone destruction in only 60% of instances. A tumor had to reach a large size to produce enough bone destruction to be radiologically recognized. Complex motion tomography allowed assessment of congenital malformations, otosclerosis, and temporal bone fractures.
Computed tomography can be performed with high-resolution bone algorithms, and a series of contiguous or overlapping CT sections can be processed to generate excellent images. Spatial resolution in MRI has increased such that information offered about the cerebellopontine angle, internal auditory canal, cochlea, and vestibule exceeds that obtained with CT. A contrast-enhanced study is of value in many situations. Magnetic resonance imaging and CT are not mutually exclusive examinations and are frequently combined to fully assess lesions of the base of the skull. Angiography is reserved for the detection of arterial stenotic lesions and arteriovenous fistulae, for the evaluation of tumor vascularity, and for preoperative embolization.
Computed tomography and MRI have become the radiologic methods of choice for most disorders of the ear and temporal bone. These high-technology studies also can be used to evaluate the neighboring skull base and related structures of the middle and posterior cranial fossae. Magnetic resonance imaging and CT are complementary. Computed tomography is good for imaging diseases that affect cortical bone, air spaces, and some soft tissue lesions of the temporal bone. Magnetic resonance imaging is excellent for evaluating soft tissues, cerebrospinal fluid, and blood vessels. However, the thin cortical bone and air-containing spaces of the middle ear and mastoid cause signal voids, which make it difficult to differentiate air from bony septations. Therefore, high-resolution CT is still the procedure of choice for bone assessment. Computed tomography is excellent for evaluating congenital anomalies, including cochlear and labyrinthine anomalies, with a special emphasis on Mondini dysplasia, congenital microtia, and other middle ear abnormalities. The degree of otomastoid pneumatization, the status of the ossicular chain, the course and position of the facial nerve, the positions of the internal carotid canal, sigmoid sinus, and jugular bulb, the thickness of the atretic plate, and the dimensions of the middle ear all are important in assessing operability. In most cases, intravenous administration of contrast material is not needed unless the study is performed to evaluate the central nervous system or vascular structures. The main disadvantages of CT are that it cannot be used to image the internal auditory canal and that it has limited value in imaging in multiple planes because many patients cannot be positioned for direct coronal or sagittal images.
In the diagnosis of erosive disease of bone, such as chronic otitis media with cholesteatoma, CT with bone window settings provides the best information. The locations, extent, and nature of the complications of cholesteatoma are ideally evaluated with CT. Acute coalescent mastoiditis, which usually occurs as masked mastoiditis, can be identified by means of thin-section axial and coronal CT examinations. Intracranial complications, such as sigmoid sinus thrombosis, extradural and subdural abscesses, and brain abscess, are best detected with contrast-enhanced CT with both bone and soft-tissue window settings. Even with MRI, dural sinus thrombosis is a difficult imaging diagnosis. Magnetic resonance angiography and MR venography are state-of-the-art examinations for evaluation of this dangerous clinical condition. Computed tomography allows the relative identification and differentiation of soft-tissue abnormalities associated with chronic ear disease, such as granulation tissue, mucopurulent effusions, cholesteatoma, and cholesterol granuloma. High-resolution CT is an excellent and sensitive means of imaging temporal bone fracture lines. Magnetic resonance imaging is recommended if the presence of coexisting intracranial abnormalities is suspected. Hemotympanum, ossicular chain disruption, and injury to the facial nerve canal are common radiologic findings. Posttraumatic cerebrospinal fluid leak is still best depicted on intrathecally enhanced CT scans.
Computed tomography with intravenous administration of contrast medium has been used routinely in screening for masses in the cerebellopontine angle. Acoustic neuroma is the most frequent tumors in this location. Computed tomography can accurately depict lesions larger than 1 cm in diameter, but it is generally unreliable for smaller tumors. Since the introduction of paramagnetic contrast agents, the usefulness of MRI has greatly improved. When a retrocochlear lesion is suspected on clinical grounds, MRI with gadolinium enhancement is the imaging modality of choice (1). It is superior to CT for evaluating acoustic neuroma because with contrast medium it can depict tumors smaller than 0.8 cm in diameter. Magnetic resonance imaging with gadolinium–diethylenetriamine pentaacetic acid contrast enhancement can help in the assessment of primary soft-tissue abnormalities in the temporal area, such as facial nerve lesions, labyrinthine schwannoma, and vestibular neuronitis. The superiority of MRI also is recognized in the evaluation of other lesions that affect the cranial nerves and cerebrospinal axis (2).
Other diagnostic modalities for vascular imaging of the temporal bone area include superselective angiography and digital subtraction angiography for screening. Angiography also can be used to direct interventional procedures. These modalities are useful in evaluating primary vascular tumors, such as glomus jugulare tumor and its extension beyond the jugular foramen into the skull base and temporal bone. These studies also are important in recognizing anomalies that can lead to surgical disasters. An aberrant middle ear internal carotid artery or high jugular bulb can be identified by means of primary vascular screening studies, by means of reconstructive or direct CT, and by means of MR angiography.
Combination radionuclide bone and gallium scans are essential in diagnosing malignant external otitis (osteomyelitis of the temporal bone) and in assessing stage and response to treatment. These scans provide physiologic information beyond the morphologic findings on CT scans.
Paranasal Sinuses
The paranasal sinuses are air-filled cavities surrounded by bone. They are inaccessible to direct clinical examination unless telescopic intervention is used. Cooperation between clinician and imager is essential for effective treatment decisions and follow-up evaluation, especially for endoscopic techniques, in which the osteomeatal complex must be thoroughly studied with CT.
Conventional plain radiography once was the imaging modality of choice in the evaluation of the paranasal sinuses. The clinical and radiographic emphasis was on disease in the maxillary and frontal sinuses. Superimposition of structures precluded accurate evaluation of the anatomic relations of the ethmoid sinuses and osteomeatal complexes. With the recognition of the importance of the osteomeatal complex in sinus disease and a change in therapeutic approach, CT has replaced conventional radiography as the primary diagnostic modality.
The target structures and views for conventional plain radiographic examination of the paranasal sinuses are summarized in Table 6.4. A complete series of sinus radiographs usually includes Waters, Caldwell, lateral, and submentovertex views; right and left oblique orbital views are obtained if the orbital apex or optic canal approximates the area of concern. Additional views may be needed to study the nasal bones or zygomatic arches. Panoramic tomography or focused dental radiographs can be used to evaluate apical dental disease that affects the maxillary sinuses. Conventional radiographs can be accurate in showing air-fluid levels, but the degree of chronic inflammatory disease present is consistently and substantially underestimated (3). Unlike conventional radiography, CT clearly depicts the fine bony anatomy of the osteomeatal complexes.
Before the introduction of CT, complex motion tomography provided critical anatomic information, especially when bone destruction or bone displacement was in question. Contiguous sections can be as close as 1 mm apart in coronal or lateral projections to detail regions of interest after a preliminary screening examination has been performed with 3- to 5-mm contiguous sections, as in assessing a blow-out fracture of the orbital floor. This technique is used if CT is not available. Although it allows clear morphologic visualization of bony abnormalities, the conventional technique has a limited gray scale.
Computed tomography has been advocated for improved diagnosis and is considered the best method for evaluating the paranasal sinuses. Imaging in the coronal plane is recommended because it optimally displays the osteomeatal units, including the relation of the brain to the ethmoid roof and the relation of the orbit to the paranasal sinuses. Coronal images closely correlate with the surgical approach used in endoscopic sinus surgery. Axial images are recommended in addition to coronal images when a patient has severe disease in the frontal, sphenoid, or posterior ethmoid sinuses, especially if surgery in these regions is being considered. The initial CT scan should be obtained after an adequate course of medical therapy to eliminate reversible mucosal inflammation (4). Modern spiral CT scanners rapidly acquire thin axial images. Use of reformatted coronal images allows examination of pediatric patients and of patients who find it difficult to maintain the position needed for coronal imaging.
Inflammatory conditions, trauma, mucocele, and tumor are critical conditions in which the use of CT is mandatory. Understanding orbital-sinus and cranial-sinus relations is essential in assessing the pretreatment stage of carcinoma of the nasal cavities or paranasal sinuses, response to radiation therapy, or postoperative tumor recurrence. Axial examinations are essential for observing breaching of the posterior wall of the maxillary sinus. Although CT can be performed without contrast enhancement, enhancement can help differentiate obstructive secretions and a mass (MRI is most useful for this purpose). Tumor extension beyond the sinuses into the orbit, brain, or retromaxillary region is best seen with contrast enhancement.
Contrast enhancement techniques are important in evaluating vascular lesions. Angiofibroma of the nasopharynx necessitates use of this technique to identify the epicenter and extension of angiofibroma from the region of the sphenopalatine foramen, medially into the nasopharynx or laterally into the infratemporal fossa and anteriorly into the maxillary sinus or superiorly into the middle cranial fossa. Cerebrospinal fluid leak and meningocele are two neurogenic disorders that can necessitate intrathecal CT enhancement. Computer-generated three-dimensional CT images are being used increasingly in evaluating complex facial fractures and severe craniofacial anomalies. Computed tomography and MRI can be used intraoperatively during image-guided sinus surgery. Conventional radiographs, complex motion tomography, and CT have common radiographic signs (Table 6.5) (5).
Magnetic resonance imaging is most useful in the evaluation of regional and intracranial complications of inflammatory sinus disease, in the assessment of benign versus malignant sinus opacification. and in the evaluation of the extent of neoplastic processes (6,7). Compared with CT, MRI provides better visualization of soft tissue, but it does not optimally display the cortical air-bone interface. Therefore, CT is still a more reliable operative “road map” for a surgeon performing an endoscopic procedure on the sinuses. Staging systems are used to evaluate CT scans of the sinuses (8).
The introduction of MRI has influenced the diagnosis, management, and follow-up evaluation of tumors of the nasal cavities, paranasal sinuses, and nasopharynx. Magnetic resonance imaging is better than CT in the assessment or characterization of soft-tissue mass lesions. Magnetic resonance imaging depicts vascular structures without the use of intravenous contrast agents. Inflammatory lesions have high signal intensity on T2-weighted images, and tumor masses tend to have low or intermediate T2-weighted signal intensity. If MRI shows a noninflammatory mass, biopsy should be strongly considered.
Other diagnostic modalities used to study the paranasal sinuses have lesser but selective importance. Ultrasonography of the maxillary and frontal sinuses has limited usefulness but can help identify fluid when conventional radiographs show opacification. Radionuclide bone scans with technetium 99m methylene diphosphonate depict osteoblastic activity in several situations, particularly identification of focal manifestations of systemic disease, such as Paget disease. Detection of these nonspecific findings is highly sensitive, and the bone scan technique shows physiologic changes before morphologic studies with conventional radiographs and CT scans do. Gallium citrate scans, if used in combination with bone scanning, allow diagnosis of osteomyelitis and follow-up assessment of the effectiveness of treatment. Treatment is considered successful when results of a gallium scan return to normal. This modality specifically images the infective focus, unlike bone scanning, which images the osteoblastic response around the infective focus (9).
Soft Tissues of the Neck
Head and neck masses have traditionally been classified as benign or malignant, primary or metastatic, and congenital or inflammatory. This classification is further extended to cover age groups, such as children and adults, and location, such as midline and lateral, anterior triangle and posterior triangle. Although the most important preliminary step in evaluating neck masses is a careful physical examination of the neck and all mucosal surfaces, this often only establishes a working diagnosis or defines an outstanding clinical problem (Table 6.1). The uncertainty associated with clinical diagnosis is determined by limitations in differentiating solid and cystic lesions and in determining anatomic associations. Radiologic imaging is essential in responding to these clinical problems.
Conventional plain radiographs are not usually helpful in differentiating neck masses, except for recognizing infrequent signs such as calcification. Ultrasonography is a safe, relatively inexpensive, and readily available investigative method. Ultrasonography is categorized as high-resolution real-time imaging or as Doppler-based technique used to assess flow characteristics of the major blood vessels of the neck. Palpable lumps in the neck can be assessed with ultrasonography, which can be used for specific assessment of growth (size of lesion over time); location; relation of lesion to the adjacent structures, especially blood vessels; character of the lesion (solid, cystic, complex); and the number and size of affected lymph nodes in the region. Applications of ultrasonography include assessment of thyroglossal duct abnormalities, branchial cleft cysts, cystic hygroma, major salivary gland tumors, inflammatory processes that progress to abscess formation, and carotid body tumors, lymph node staging, and imaging-guided FNA.
Ultrasound techniques combined with FNA and cytologic evaluation have particular importance in the assessment of the soft tissue of the neck (10). The superiority of ultrasonography is only in its nonradiographic guidance capability including FNA biopsy of cervical lymph nodes or other masses. Ultrasonography also can be useful for percutaneous drainage of cervical abscesses. Fine needle aspiration biopsy guided by computed tomography can be used to diagnose poorly accessible or deep-seated lesions of the head and neck (11).
Computed tomography and MRI are widely used for primary staging of tumors and nodes. However, accuracy in assessing lymph nodes depends on the radiologic criteria used. Various criteria for the size of lymph nodes have been used for diagnosing metastatic lymphadenopathy, but the most reliable imaging finding is the presence of nodal necrosis. Areas of central nodal necrosis larger than 3 mm are routinely identified on contrast-enhanced CT scans. The usefulness of MRI and CT in the detection of lymph node metastasis from squamous cell carcinoma of the head and neck was assessed by Curtin et al. (12). Computed tomography performed slightly better than MRI in the detection of lymph node metastasis. A high negative predictive value was achieved only when a low size criterion (5 to 10 mm) was used and was therefore associated with a relatively low positive predictive value. For example, with criteria of 1-cm size or an internal abnormality to indicate a positive node, CT had a negative predictive value of 84% and a positive predictive value of 50%. Magnetic esonance imaging had a negative predictive value of 79% and a positive predictive value of 52%. Ultrasonography is hampered by similar morphologic criteria, and only ultrasound-guided FNA biopsy can offer additional cytologic criteria that may be more reliable (13). Positron emission tomography with 18-fluorodeoxyglucose in the detection of cervical lymph node metastasis has been shown in several studies (14,15) to have a higher sensitivity and specificity than CT, MRI, or ultrasound-guided FNA biopsy.
Compared with MRI, CT is the best method for studying capsular penetration and extracapsular nodal extension. With contrast CT, extracapsular extension is identified with an enhancing nodal rim, usually with infiltration of the adjacent fat planes. Carotid artery invasion is an important prognostic indicator, and invasion of the adventitia is as important as greater degrees of arterial invasion. Detection of microscopic adventitial infiltration is beyond the scope of current imaging, but circumferential involvement of the artery by tumor is strong evidence that the artery is invaded.
Both CT and MRI are useful to evaluate for tumor recurrence and posttreatment changes in the care of patients with malignant tumors of the head and neck. Magnetic resonance imaging has advantages in differentiation between tumor and scar tissue, but edema after radiation therapy can make differentiation difficult. Radiation-induced changes lead to false-positive diagnoses for approximately 50% of patients (16). Criteria for recurrent or residual tumor include an infiltrative mass with high signal intensity on T2-weighted images and enhancement after administration of gadolinium on T1-weighted images. Computed tomographic criteria for recurrent or residual tumor include the combination of a circumscribed, infiltrative mass with contrast enhancement on CT scans. Positron emission tomography compares favorably with CT and MRI in the detection of recurrent or residual cancer (17).
Other methods for detecting tumor recurrence include thallium 201 single photon emission computed tomography (SPECT) and gallium 67 citrate whole-body scintigraphy. Mukherji et al. (1
found thallium 201 SPECT to be superior to CT for differentiating recurrent tumor from posttreatment changes. Murata et al. (19) found gallium 67 citrate whole-body scintigraphy especially useful for evaluation for recurrence and distant metastasis of squamous cell carcinoma of the head and neck.Computed tomography is useful for detecting clinically occult primary cancer of the head and neck that causes cervical metastasis. Magnetic resonance imaging can better depict submucosal lesions in the areas of the tonsils, anterior floor of mouth, and base of the tongue. Magnetic resonance imaging should be performed first for a mass in the retropharyngeal or parapharyngeal space. In these locations, tumors of neural, vascular, or salivary origin should be suspected. Magnetic resonance imaging also should be performed if posterior extension into the airway, esophagus, or posterior deep muscles is suspected. With few exceptions, CT is better than MRI in depicting thyroglossal duct cysts, branchial cleft cysts, and cystic hygroma and in differentiating infection, inflammatory processes such as cellulitis, edema, and abscess.
Digital subtraction angiography and conventional superselective angiography are useful in the diagnosis of hemangioma, arteriovenous malformation, and paraganglioma. Intra-arterial embolization has important therapeutic applications as definitive treatment or before surgery.
Despite being hypovascular, nerve sheath tumors can become greatly enhanced on CT scans, apparently because of extravascular leakage of contrast material into the tumor bed. Nerve sheath tumors have intermediate signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Paraganglioma becomes intensely enhanced after intravenous administration of contrast material during CT and MRI studies. The characteristic salt-and-pepper appearance on MR images reflects the signal voids of many tumor vessels. However, signal voids do not occur with all paragangliomas, especially in those less than 2 cm in diameter.
Larynx
Radiologic imaging modalities are important in screening and in defining the deep dimension of a malignant tumor of the larynx. Although the larynx is readily accessible to direct visualization, including telescopic assessment, and biopsy, submucosal extension is not subject to direct visualization. If possible, imaging studies should be performed before biopsy to avoid confusion of tumor and local trauma. Vocal cord mobility defects, whether due to direct infiltration by tumor or involvement of the recurrent laryngeal nerve, can be assessed with diagnostic imaging. There also are difficult-to-evaluate regions, such as the Morgagni ventricle or the subglottis. Imaging is critical in the evaluation of carcinoma of the larynx in all three major regions (supraglottis, glottis, subglottis) and in evaluation of extralaryngeal extension of malignant growth to the hypopharynx or the laryngeal cartilage. Virtual endoscopy of the airway is possible, but it has limited usefulness in differentiating mucosal surfaces that are touching (20).
Conventional plain radiographs (lateral and anteroposterior projections and selective high-kilovoltage filtration techniques) of the larynx provide preliminary or definitive information about foreign bodies, trauma, and other types of acute and chronic airway obstruction. These radiographs can show soft-tissue swelling, alteration of the cartilaginous framework if sufficiently calcified, and the position of the air column. The variability of calcification of the laryngeal cartilage can pose a diagnostic problem in the detection of foreign bodies. For example, the superior margin of the cricoid cartilage calcifies long before the signet portion does. The result is linear calcification on plain radiographs that often is mistaken for a foreign body.
Xeroradiography, because it provides edge enhancement, can clarify intrinsic soft-tissue detail such as calcification, delineate masses and stenosis, sometimes depict cartilage abnormalities such as fractures and erosions, and help identify foreign bodies by type and location. Unfortunately, this technique carries a radiation exposure three to five times that of conventional radiography. The usefulness of xeroradiography in imaging of the soft tissue and cartilage has been superseded by that of CT and MRI.
Conventional coronal tomography allows visualization of frontal view anatomy without a superimposed spine. Thus it allows satisfactory analysis of the vertical extent of laryngeal tumor and subglottic or tracheal stenosis or stricture. This technique has been replaced by CT and MRI because of its limited gray scale for soft-tissue differentiation, but the airway image, especially with the added sagittal projection, is excellent. Tomography has several limitations, such as poor definition of the anterior commissure. Depiction of cartilage invasion is unreliable, except when there is extensive involvement of well-calcified cartilage.
Ultrasonography has limitations because the laryngeal cartilage reflects most of the sound, limiting ultrasonic access. Nuclear medicine imaging seldom is useful for laryngeal imaging. Increased uptake has been mentioned as a cause of inflammatory arthropathy and relapsing polychondritis. Erythrocyte-tagged imaging can provide enough information for diagnosis of the rare laryngeal cavernous hemangioma. Arteriography is seldom used except for evaluation of a suspected vascular lesion such as paraganglioma.
Technologic advances in CT and MRI have greatly improved the ability to image the larynx. Spiral CT and fast MRI techniques allow rapid acquisition, which decrease degradation motion artifacts from breathing, swallowing, and carotid artery pulsation. Both CT and MRI allow evaluation of the extent of laryngeal tumors, especially for tumor size staging of carcinoma. Such determinations can influence the extent of laryngectomy (partial versus total).
Spiral CT scanners acquire the complete data set through the larynx in less than 10 seconds, allowing the patient to stay motionless. Images can then be reconstructed to give overlapping sections, and coronal, sagittal, and even three-dimensional images can be generated from the same data set. Examination of the larynx during various inspiratory-expiratory cycles has been used to optimize visualization of a particular region or the margin of a tumor. With operator-directed imaging, CT provides accurate images of the location, size, and extent of the tumor. Computed tomography can depict cartilage invasion and deep soft-tissue spread into the paraglottic space, preepiglottic space, and pyriform sinuses. This T-stage imaging contributes to a more rational diagnosis and more effective planning of surgical and radiation treatment.
The larynx has been difficult to image well with MRI because of motion artifact. Fast spin echo imaging has made a marked difference in the ability of MRI to image the larynx by reducing these motion artifacts. The use of gadolinium is controversial. Some experts believe gadolinium enhancement provides important information regarding the interface of tumor with muscle. However, fast spin echo imaging can generate similar information, obviating administration of gadolinium. Magnetic resonance imaging is better than CT at separating soft tissues. Another advantage of MRI over CT is acquisition of high-resolution images in multiple planes. Sagittal images show the epiglottis, vallecula, and base of the tongue well. Coronal views are ideal for evaluating the margins of the vocal cords, the paraglottic space, and the vertical extent of tumor. Axial images allow assessment of cartilaginous erosion.
The appearance of the cartilage on CT scans and MR images varies with the degree of ossification, which is not uniform and frequently is asymmetric. Invasion of cartilage has implications in staging and outcome of carcinoma of the larynx. With CT, tumors often have the same attenuation as nonossified cartilage (soft-tissue density), so minimal cartilage involvement can be difficult to assess. Gross cartilage destruction with tumor on the opposite side of the cartilage from the primary lesion is the only truly reliable CT sign of cartilaginous invasion. This finding is similar in MRI, but there is evidence to suggest that lesser degrees of tumor involvement can be detected with MRI because of variability in the appearance of both normal and abnormal cartilage afforded by MRI sequences. Nodal staging can be dramatically improved with sensitive broadening of the primary CT examination. Thin-section CT added to the primary tumor assessment can show enlarged or metastatic nodes in the neck and specific anatomic and pathologic features.
Salivary Glands
The diagnosis of salivary gland disorders is established from the findings of the history, physical examination, and FNA biopsy rather than from radiographic studies. However, imaging studies often are needed to assess focal, multifocal, diffuse, and bilateral disorders. New diagnoses, such as stones, can be readily established. Imaging studies can be used to confirm a clinical suspicion, such as human immunodeficiency virus–positive status from parotid gland enlargement with lymphoepithelial cysts.
Conventional radiographic examination that includes occlusal radiography and panoramic tomography can depict radiopaque duct calculi. Most submandibular duct and gland stones are radiopaque. Parotid duct stones occur infrequently and usually are radiolucent. Conventional sialography once was the standard method of assessment of the morphologic features of the salivary duct and glands. It is not commonly performed now, although it provides important information about nonopaque stones, ductal stenosis, sialectasis, and sialosis.
In imaging of the major salivary glands, ultrasonography has reasonable accuracy in differentiating intracapsular and extracapsular lesions. It also images solid and cystic lesions. Mixed solid-cystic lesions, such as Warthin tumor, are considered solid. Ultrasonography is limited in depicting deep-lobe parotid lesions because of the attenuation and reflection of sound by the mandible.
Radionuclide scanning is useful in some pathophysiologic salivary gland assessments because 99mTc-pertechnetate is incorporated into the salivary glands and excreted in the saliva. Increased focal uptake is characteristic of functioning tumors, such as Warthin tumor of the parotid gland and the rare oncocytoma. Diffuse, increased radionuclide uptake often indicates ductal obstruction leading to intra-acinous salivary retention in the presence of long-standing inflammation. Positron emission tomography does not reliably differentiate benign from malignant tumors, limiting its clinical usefulness. It also is expensive and takes a long time to perform.
Computed tomography and MRI are excellent methods for imaging and evaluating salivary gland disease, specifically focal, multifocal, and diffuse masses. The two modalities have equivalent diagnostic potential for imaging solid and cystic lesions. Both provide important information about the location (intraglandular or extraglandular), size, and extension of tumor to surrounding superficial and deep structures.
Bone invasion is viewed differently with MRI and CT, and marrow invasion of the mandible is better identified with MRI. The excellent soft-tissue depiction with MRI allows identification of the intraparotid portion of the facial nerve and its relation to infiltrative masses, which allows planning of surgical management. Magnetic resonance imaging also is superior to CT in evaluating the muscle-tumor interface. Another advantage of MRI over CT is the absence of exposure to radiation or the intravenous administration of iodine-containing contrast medium.
The choice of which imaging study to perform in investigating salivary gland disease is influenced by clinical presentation, user preference, and familiarity with a specific modality. Computed tomography is the choice in cases of inflammatory disease of the salivary glands, because MRI does not depict ductal dilatation or salivary calcification. However, if the clinical finding is a mass, the initial imaging evaluation, if any, usually is MRI. Computed tomography is an acceptable alternative, and ultrasonography can be used as a complementary study.
Thyroid Gland
The superficial position of the thyroid gland in the neck enables easy access for clinical examination and FNA. A wide range of high-technology imaging modalities are available for the diagnosis and management of thyroid disease. These may generate structural information, as do ultrasonography, CT, and MRI, or show tissue function, as does radionuclide scanning (21).
Conventional radiography is not a primary study. It is limited to screening for airway or esophageal displacement or invasion and to identification of calcification in the thyroid gland. Radionuclide scanning often is chosen for imaging malignant lesions of the thyroid. Three radioactive pharmaceuticals are commonly used in clinical practice. Sodium 99mTc-pertechnetate is trapped by the thyroid gland but is not organified. Radioiodine isotopes (iodine 123 and iodine 131) are trapped and organified by the parenchyma of the thyroid gland. Pertechnetate is the most commonly used radioisotope, at least for the initial evaluation. It is less expensive than radioisotopic iodine, is readily available, and approximates iodine trapping through the metabolism of the pertechnetate anion, which is not incorporated into hormonogenesis. Because 123I radioiodine scanning is unique in providing anatomic images and images of the functional activity of the thyroid gland or ectopic thyroid tissue, it is useful in a variety of clinical situations. These include investigation of a palpable thyroid nodule or a mass in the midline of the neck, base of the tongue, or mediastinum. Radiopharmaceuticals can be used in the management of cancer and Graves disease and in screening for thyroid metastasis and postsurgical recurrent tumor.
Radionuclide scanning has several limitations. The anatomic imaging resolution is only 1.0 cm, which restricts detail and definition. Scans may not give an adequate image of thyroid tissue if the patient is taking oral thyroid hormone supplements. A hot nodule on a sodium 99mTc-pertechnetate scan may or may not indicate a functioning tumor. In these situations, a radioactive iodine (123I) scan shows physiologic hormonogenesis and a functioning or nonfunctioning mass.
High-resolution ultrasonography is the first-line structural investigative modality in the diagnosis of many thyroid disorders, especially nodular disease. It is safe, inexpensive, simple, quick, and reproducible. Marked improvement in image quality occurred with the introduction of small-parts ultrasonography. Ultrasonography has an accuracy greater than 90% in differentiating cystic and solid thyroid nodules. Mixed solid-cystic nodules should be managed as solid masses. Ultrasonography can depict questionable or difficult-to-palpate lesions and can be used to direct FNA. It also can show whether a palpable nodule is part of a focal, multifocal (multinodular goiter), or diffuse process. Ultrasonography can be used to follow the size of a nodule after suppression therapy or cyst aspiration. The major limitation of ultrasonography is the inability to differentiate malignant from benign lesions on the basis of tissue characteristics. This is appropriately left to a pathologist. Retrosternal thyroid cannot be evaluated because of the bony interference with sound.
Computed tomography and MRI have similar roles in the evaluation of thyroid disorders and are essentially second-line imaging modalities for this region of the neck. Both techniques provide useful information about the size, shape, and anatomic structure of thyroid nodules. They can help determine whether a mass is solitary or part of a multinodular lesion. Computed tomography and MRI can be used to evaluate mediastinal, substernal, or retrosternal extension of thyroid masses and regional lymph node involvement or local recurrence. Computed tomography shows calcification better than does MRI, but MRI is superior in providing soft-tissue detail, especially for the muscle-tumor interface. Magnetic resonance imaging can be performed without intravenous contrast medium and does not expose the patient to radiation. If iodinated contrast material is used during CT, the iodine interferes for months with thyroid function tests and uptake of iodine 131 used postoperatively to manage well-differentiated malignant disease of the thyroid.
Parathyroid Glands
Parathyroid imaging is controversial in terms of the indications for imaging and the imaging agent used. Preoperative imaging localization studies are not performed at many centers if the patient has not undergone surgical intervention (22). There are many options for imaging the parathyroid glands, including ultrasonography, CT, MRI, angiography and many nuclear medicine studies. Ultrasonography is the least invasive. Because of their small size, normal parathyroid glands are not usually detected with ultrasonography. However, parathyroid lesions larger than 0.5 cm in diameter usually can be identified in a careful study. In cases of hyperparathyroidism due to parathyroid adenoma, accurate localization is reported at a rate between 69% and 88% (23). Our success rate is greater than 90%. For perithyroidal parathyroid adenoma, ultrasonography is an excellent imaging choice, but acoustic impedance prevents adequate imaging behind air-filled structures (trachea) or bone (mediastinum).
No nuclear medicine agent is exclusively taken up by normal or adenomatous parathyroid glands. The agents that are taken up by both parathyroid adenoma and thyroid tissue (thallium 201 and 99mTc-sestamibi) are subtracted from agents that are taken up by thyroid tissue only (99mTc-pertechnetate and iodine 123). However, 99mTc-sestamibi imaging is commonly performed without subtraction because this agent washes out of the thyroid gland rapidly but is retained by parathyroid tissue and thyroid adenoma. Although these scans have limited value in detecting four-gland hyperplasia but have some usefulness in detecting asymmetric hyperplasia, diagnostic accuracy is reported at between 50% and 95% in cases of parathyroid adenoma (24).
Digital subtraction angiography relies on a single morphologic characteristic—the regional blood supply to a hypervascular gland. After the intra-arterial or intravenous injection of contrast medium, a computerized image-subtraction program eliminates unnecessary background information. This technique allows identification and localization of parathyroid adenoma in 60% to 70% of patients. This modality rarely is needed because sestamibi scans and ultrasonography are so successful. It may be useful in revision cases of hyperparathyroidism.
Normal parathyroid glands seldom are depicted on CT scans. Markedly enlarged glands can be detected with CT with a sensitivity of 50% to 88%. In the neck, an axial CT image is no more informative than ultrasonography. Computed tomography is limited by technical factors, such as inadequate resolution of lesions less than 1.0 cm in diameter, or by limitations in interpretation, such as mistaking a tortuous vessel, thyroid mass, or lymph node for an enlarged parathyroid gland. Use of iodinated contrast agents for differentiating blood vessels from adenoma or lymphadenopathy prevents subsequent imaging with iodine-based nuclear medicine studies for approximately 6 weeks.
Normal parathyroid glands usually are not identified with MRI. Magnetic resonance imaging with gadolinium enhancement can be useful for evaluating hyperparathyroidism refractory to the first surgical attempt at correction. T2-weighted MRI and gadolinium fat-suppressed T1-weighted scans show parathyroid adenoma as an area of high signal intensity against a dark, soft-tissue background. A review of the parathyroid literature up to 1993 showed that MRI had the highest sensitivity for the detection of adenoma (74%) followed by nuclear medicine studies (72%), CT (65%), and ultrasonography (63%) (25). At reoperation for previously unidentified adenoma, MRI had the highest (66%). Ultrasonography had a sensitivity of 60%; CT, 48%; and nuclear medicine studies, 45%.
0 comments:
Post a Comment