Diagnostic imaging in salivary gland disease

radioiogy Ec itor. JOHN

W. PRE ECE, DD ;

American Academy 0,. Der tal Radiology Department of Dental Dia;:nostic Sciences SChool of Dt ntistry, l’he cbziversity of Texas Health Science Center at San Antonio 7703 Floyd Curl Dr. San Antonio Texas 75284

Diagnostk imaging in salivary gland disease Ham P. van tien Akker. DDS, PhD,* Amsterdam, UNIVERSITY

The Netherlands

OF AMSrERDAM

A review of the zurrelt status of salivary gland imaging is presented. The potentials and limitations of sialography, sci itigraahy, ultrasonography, computed tomography, and magnetic resonance imaging are discussed so that a rational approach to their use can be adopted. (ORAI. SURC ORAI MED ORAL PATHOL 1988;66:625-37)

re imaging procedures usually necessary in the A diagnostic workup of :,alivs.ry gland diseases,and, if sol,what are the indications for the various methods? The marked increase in imaging procedures in the last decade has made these questions even more relevant today than they w’ere in the past. Clinical signs and syml>toms of diseases of the salivary glands generally tend to be nonspecific. However, the superficial location of these glands easily permits inspection and palpation so that a thorough physical examination combined with a detailed history often Iprovidessufficient information for a clinical diagnos,is. Nevertheless, confirmation of this diagnosis and further evaluation are frequently necessary,and imaging procedures play an important role in this respect. When further investigations have to be conducted, lit is apparent that the clinician not only should have a knowledge of what methods are available but also should appreciate the potential value and limitations of each method so that he can make a rational choioe. The purpose of this article is therefore to discuss the current status of the various imaging procedures and to present a survey of their application in salivary gland disease (Table I). SIALOGRAPHY

Sialography still is the most widely used radiographic method for examination of the salivary glands. Nevertheless, Halt.’ once characterized sia*Department of Oral and Maxillofacial Surgery, Academic Medical Center and Academic (Center for Dentistry.

lography as the neglected stepchild of roentgenographic diagnosis, becausethroughout the years this diagnostic procedure has remained controversial; being considered worthless by some authors and indispensable by others. However, the question of whether sialography is useful cannot be answered in the absolute sense, because the answer largely depends on the type of salivary gland disorder under investigation. Overrating the diagnostic information that can be obtained may lead to indiscriminate use of sialography and subsequently to disappointing results, particularly in tumor diagnosis, and this is probably the main reason why many clinicians are dissatisfied with the method. Furthermore, since negative experiences may reflect unfavorably on sialography as a whole, the danger exists that its use will be discarded altogether, even in cases in which useful information can be obtained. It is therefore extremely important to appreciate fully the true indications for sialography so that unrewarding examinations can be avoided as much as possible. Most authors nowadays agree that sialography is of limited use in tumor diagnosis.2-4Space-occupying lesions are only indirectly visualized by displacement or disruption of the ductal system, and even though more than 80% of tumors can be detected by a refined sialographic technique, lesions less than 1 cm in diameter usually are not demonstrated.5,6Furthermore, differentiating between an intrinsic and extrinsic location and predicting the benign or malignant nature of a tumor are not always possible.‘,* Nevertheless, sialography still retained some popularity as 625

626

van den Akker

Oral Surg November1988

Fig. 1. A, Parotid gland sialogram of a patient with Sjogren’s syndrome shows globular collections of contrast material evenly distributed throughout the gland. B, The scintigraphic image reveals decreasedand asymmetric pertechnetate uptake in the parotid glands and absenceof uptake in the submandibular glands. C, Parotid gland sialogram of a patient with sialosis. The ducts are thin and slightly separated but otherwise normal. D, The scintigraphic image reveals markedly increased pertechnetate uptake in the enlarged parotid and submandibular glands.

Table

I. Usefulness of imaging procedures in salivary gland disease

I Imaging procedure

Sialography Scintigraphy Ultrasonography CT MRI Plain radiography

Type of disease Tumors and cysts

Sialoadenitis

* t

Sjcgren’s syndrome

Sialolithiasis and obstructive disorders

*

* *

r

5

*; * *

Sialosis

* *

Congenital anomalies and traumatic disorders

* * t t

*May provide useful information. fUseful in selected casesonly. *tIf necessary, in combination with sialography or intravenous contrast medium infusion.

a method for evaluating space-occupying lesions until the introduction of computed tomography (CT), and more recently magnetic resonance imaging (MRI), finally made the question of its usefulness as a sole procedure irrelevant (vide infra).

The main indications for sialography are inflammatory and nonneoplastic diseases,which generally show changes in the ductal system itself. These disorders form a heterogeneous group consisting of the various types of sialoadenitis, including Sjiigren’s

Volume 66 Number 5

Diagnostic imaging in salivary gland disease

627

Fig. 3. Persistent accumulation of pertechnetate in the lower pole of the left parotid gland after stimulated excretion (arrow): Warthin’s tumor.

Fig. 2. Right parotid gland sialogram shows a calculus expressing itself as an oval filling defect (arrow). This calculus was not visible on plain radiographs.

syndrome, granulomatous diseases, obstructive disorders such as sialolithiasis and strictures, sialosis, traumatic lesions, and developmental anomalies. In many instances their clinical pictures are similar, with only a limited number of nonspecific signs and symptoms such as pain, persistent or recurrent diffuse swelling, or diminished function. However, their sialographic images exhibit a wide variety of ductal changes that are often sufficiently characteristic for the clinician to differentiate one disease from the other.9 For instance, both SjGgren’s syndrome and sialosis may appear with painless, often bilateral, and persistent parotid gland enlargement and xerostomia so that clinically the differential diagnosis may be difficult. In Sjiigren’s syndrome, however, sialography will generally show sialectasias distributed throughout the parotid gland with or without dilation, of the main ducts, whereas in the case of sialosis the ductal system of the enlarged gland appears essentially normal even though the sparse peripheral branches are more separated and narrower than usual because of compression by the hypertrophied serous acinar cells. If a contraindication for sialography exists, differentiation between Sjiigren’s syndrome and sialosis is also possible by scintigraphy with 99mTc-pertechnetate, which may show a decreasedisotope uptake in Sjogren’s syndrome or a characteristic increased isotope concentration in sialosis’” (Fig. 1). Clinical signs and symptoms of salivary gland obstruction are often pathognomonic, but nevertheless further evaluation frequently requires a sialo-

graphic examination. Although radiopaque calculi can usually be detected on preliminary plain radiographs, approximately 20% of submandibular gland and 40% to 70% of parotid gland calculi are reported to be poorly calcified and therefore cannot be demonstrated by plain radiography.“,‘: However, on a sialogram these radiolucent stones may be visualized as a filling defect or, more rarely, as an area of narrowing of the duct with proximal dilatation (Fig. 2). Whether these stones are truly radiolucent, or in fact radiopaque but merely difficult to visualize with plain radiography remains the question. The latter view is supported by Suleiman and Hobsley,13who, by using lateral, anteroposterior, and intrabuccal plain radiographs, were able to demonstrate more than 70% of parotid gland calculi.13 Sialography usually is not indicated when a radiopaque stone is situated in the distal part of the duct, because manipulation with the catheter or pressure during injection of the contrast medium may displace the stone deeper into the duct, possibly even to a location inaccessible from the mouth.14However, this objection does not apply to deeply situated stones. In these cases sialography provides valuable information on the stone’s location, which determines the operative procedure necessaryfor its removal. When a stone is situated in the intraglandular part of the duct, glandular excision or removal of the stone via a parotidectomy approach is indicated, whereas in the case of an extraglandular location transoral removal is the treatment of choice.‘ss’6The extent of glandular damage, as seenon a sialogram, is not the decisive factor in the question of whether a gland should be excised or left in situ, because of the remarkable capacity for functional recovery once the obstruction is removed.16 Occasionally confusion may be caused by the presence of phleboliths, calcified lymph nodes, or other paraoral pathologic calcifications. These radiopaque structures are normally readily distinguishable from salivary calculi, but in case of doubt a

62%

van den Akker

Fig. 4. A, Scintigraphic image of a patient with a stone in the knee of the right submandibular duct shows absence of pertechnetate uptake in the right submandibular gland (urrow). B, Three years after transoral removal of the stone there is normal pertechnetate uptake in this gland (arrow).

sialographic study will demonstrate their location outside the ductal system.“.‘* Finally, as a morphologic study sialography is also useful for detecting and evaluating strictures, salivary fistulas, sialoceles, and developmental anomalies,g,‘gbut for technical reasons the procedure cannot always be carried out. The alternative is then to perform salivary gland scintigraphy with g9mTcpertechnetate.*O SCINTIGRAPHY

WITH 9smT~-PERTECHNETATE

In contrast to other imaging procedures scintigraphy is primarily a function test, although the method also provides some information on the structure and topography of the major salivary glands. The examination is carried out with a gamma camera after intravenous injection of 74 MBq g9mTc-pertechnetate, which is selectively concentrated and secreted by the salivary glands. Sequential analog images, usually made during a period of 60 to 90 minutes, permit a continuous visual monitoring of the salivary dynamics with respect to pertechnetate. However, acceleration of pertechnetate excretion by oral administration of citric acid or subcutaneous injection of 0.25 mg of the parasympathomimetic agent carbachol (carbamylcholine chloride) at 10 or 20 minutes after pertechnetate injection shortens the examination time to a more acceptable 30 or 40 minutes and, moreover, provides a rapid means of investigating ductal patency.*’ Data may be simultaneously digitized with an on-line computer system and later replayed to obtain time-activity curves from regions of interest placed over the salivary glands and background areas over the skull. Ratios calculated from these curves or absolute uptake

measurements are useful to quantitate glandular function in selected cases2’.?* In the diagnosis of salivary gland tumors the results of scintigraphy have been disappointing.?‘-” The resolution of the imaging system is such that only tumors exceeding 1.5 cm in diameter arc visualized. Furthermore, differentiation between benign and malignant massesis not possible, because both types generally do not concentrate pertechnetate, which results in “cold spots” on the scintigraphic images. The only exceptions are Warthin’s tumor and oncocytoma, which generally take up more pertechnetate than the surrounding normal glandular tissue and thereby appear as “hot spots” (Fig. 3). In the wider area of head and neck oncology, scintigraphy has successfully been used to study the effect of radiation on salivary gland function2hand to detect subclinical salivary leakage after extensive neck surgery as a major cause of postoperative wound infection.27 Scintigraphic results in salivary gland infections depend on the activity of the inflammatory process. Acute sialoadenitis usually shows diffusely increased pertechnetate concentration in the enlarged gland as a result of hyperemia and delayed or absent isotope excretion caused by compression of the ductal system by inflammatory edema. Chronic sialoadenitis, on the other hand, frequently results in a loss of ability to concentrate pertechnetate, which reflects the parenchymal damage and interstitial fibrosis. Pertechnetate uptake in the diseased gland may range from almost normal to absent, depending on the severity and duration of the inflammatory process. However, most salivary gland infections can easily be diagnosed on the basis of the history and clinical observations so that scintigraphy is generally not necessary for the diagnosis per se. Scintigraphy at a later stage may be useful, especially in the case 01 recurrent infection, to assessthe rate of recovery as a guideline for further surgical or medical treatment.28,29Furthermore, the method may show that more glands are involved than are clinically evident. Several authors suggest that in patients with sialolithiasis, scintigraphy can be used as a preoperative test to determine the type of treatment: Markedly decreasedor absent pertechnetate uptake in the involved gland would indicate the necessity for glandular excision, whereas normal or only slightly decreased pertechnetate uptake would indicate that surgical removal of the stone, leaving the gland in situ, should be attempted.21-29-“’However, recent observations have shown that markedly decreasedor

Volume 66 Number 5

Fig. 5. Ultrasonographic image of Warthin’s tumor shows sharply defined lateral and posterior margins and some internal echoes. There is echo enhancement behind the tumor.

absent pertechnetate uptake preoperatively does not necessarily rule out the possibility of functional recovery after stone removaP (Fig. 4). The predictive value of scintigraphy in these casesis therefore limited and makes the method unsuitable as a routine test in the preoperative work-up. As a noninvasive procedure scintigraphy is particularly useful in the diagnosis and follow-up of patients with Sjiigren’s syndrome or other systemic diseases involving the salivary glands (Fig. 1). The main advantage of scintigraphy in these instances is the possibility of examining all major salivary glands simultaneously and continuously over a period of time so that the full extent of glandular dysfunction can objectively be demonstrated. In Sjiigren’s syndrome the results of scintigraphy correlate well with the patient’s subjective symptom of xerostomia and with the results of other diagnostic methods, such as sialography, sialometry, and minor salivary gland biopsy.32-34For this reason some authors do not perform sialography or sialometry but instead use scintigraphy, along with a parotid or labial salivary

Diagnostic imaging in salivary gland disease

629

Fig. 6. Sialolithiasis of right parotid gland. On the ultrasonographic image the calculus is visible as an echodense spot (arrow) casting an acoustic shadow. The parotid duct was not accessible after several attempts to remove the calculus via an intraoral approach.

gland biopsy, to diagnose and quantitate the oral component of Sjiigren’s syndrome.34 An absolute indication for scintigraphy exists when the ductal orifice of a major salivary gland cannot be found or cannot be cannulated.‘O This problem occurs when, for no obvious reason, one or several ductal orifices are not present in their usual location; this condition suggests the possibility of glandular aplasia. The problem may also be due to the presence of scar tissue or to an inflammatory infiltrate in the vicinity of an excretory duct that results in transpositioning or stenosis of the distal part of the duct with or without a fistulous redirection of the salivary flow. Furthermore the main excretory duct is usually not accessibleafter surgical procedures such as glandular excision, ductal ligation, or ductal repositioning toward the oropharynx as treatment for drooling. After these operations signs or symptoms may persist or develop that require further evaluation, or it may be considered of interest to objectively document the effect on glandu-

630

van den Akker

CM November

Fig. 7. Axial CT scan through the middle normal parotid gland. The gland is distinctly dense than adjacent fat in the subcutaneous parapharyngeal space but less dense than muscles.

Surg 1988

portion of a more radiotissues and surrounding

lar function of a clinically successful ductal ligation or repositioning. Finally, in some patients an attempt at sialography may fail, because it proves to be technically impossible to probe and cannulate an excretory duct, particularly the submandibular duct, even though the orifice is patent and in its normal position. It is apparent that in all these instances salivary gland scintigraphy is the only diagnostic procedure available to confirm the clinical diagnosis and to assessglandular function, if present, including the residual patency of the main excretory duct. An application of salivary gland scintigraphy not related to primary glandular diseaseis the use of the procedure as an alternative to the submandibular salivary flow test for early prognosis of Bell’s palsy. A scintigraphic examination performed within 10 days after onset of symptoms and, if necessary, repeated 3 weeks later will show whether the stimulated excretion of pertechnetate from the submandibular gland on the affected side is normal, decreased, or even absent in comparison with the healthy side. A recent study indicates that this parameter reliably predicts the prognosis of facial nerve paralysis and thus provides the information necessary for a choice of further treatment.35 ULTRASONOGRAPHY

Diagnostic use of ultrasonography is based on the transmission of an ultrasonic beam into the area to be scanned and the reflection or echo of the sound waves that occurs when these waves reach an interface between tissues of different density or impedance. The potential of ultrasonographic procedures

Fig. 8. Axial CT scan shows a pleomorphic adenoma of the left submandibular gland (nrrows). The gland is markedly enlarged in comparison with the normal right submandibular gland (S).

in various fields of medicine is generally acknowledged, but the value of the method in detecting salivary gland pathosis often appears to be overlooked even though the first reports on the use of ultrasound in parotid gland disorders were published a decade ago.36,37 In comparison with other imaging procedures such as sialography and CT, ultrasonography has somedistinct advantages: It is inexpensive, widely available, easy to perform, painless, and, most important, noninvasive, without any known harmful effects on the patient. Furthermore the tomographic plane can be easily manipulated at the time of imaging, and thereby the amount of diagnostic information obtained can be increased. The main application of ultrasonography in the diagnosis of major salivary gland disease is the differentiation of cystic or fluid-containing lesions from solid masses.38Cystic lesions generally contain no tissue interfaces to produce internal echoes and therefore appear as echo-free areas with the typical sign of enhancement of the posterior wall. This latter phenomenon is due to the fact that becausefluid is a good transmitter of ultrasound waves, the interface with the solid posterior wall produces a strong echo behind the cyst. Solid tumors larger than 5 mm are readily visualized by ultrasonography because they

Volume 66 Number 5

Diagnostic imaging in salivary gland disease

631

Fig. 9. Axial CT scans demonstrate the varied density of pleomorphic adenomas. A, Sharply marginated high-density lesion in the superficial lobe of the right parotid gland (arrow). B, Large lesion in the right parotid gland. The thin lateral rim of normal parotid tissue (arrows) is more radiodense than the tumor and contains some intraductal contrast material (white spots).

are less echogenic than the surrounding normal salivary parenchyma (Fig. 5). The sensitivity of the method in detecting superficial lobe parotid gland masses approaches lOO%, and tumors within the parotid gland can generally be differentiated from superficially located extraglandular lesions.39However, since bone reflects rather than transmits sound, tumors in the deep lobe of the parotid gland are mostly obscured by the mandible and the mastoid process.38Therefore CT remains the method of choice for evaluating deep lobe parotid gland masses and lesions of the parapharyngeal space. Several authors have reported ultrasonography to be helpful for differentiating benign from malignant tumors, the best sonographic criteria being the margins and the echo structure of the examined lesions.39-42 Benign tumors usually are weakly echogenic, with homogeneous echo strength and density and, most important, sharply defined, sometimes lobulated margins. Malignant tumors, on the other hand, show a nonhomogeneous echo pattern and irregular, illdefined margins, although well-defined margins have also been reported, particularly in mucoepidermoid tumors.42x43 On the basis of these criteria Wittich and co11eagues39 and Bruneton and coauthors40were able to assessthe benign or malignant nature of a lesion in 87, or 80%, of their patients. However, it should be realized that even though the sonographic pattern is sometimes sufficiently characteristic to suggest a specific tumor type, ultrasonography is not capable of providing a definite diagnosis. Therefore, aspiration cytology or excisional biopsy is always necessary to establish the histologic characteristics of a lesion. The role of ultrasonography in inflammatory and

obstructive diseases is limited, sialography usually being more informative in these cases. However, it may be difficult to clinically differentiate acute suppurative parotitis from an acute abscess,and for obvious reasons sialography is not indicated in these conditions. Therefore in case of doubt ultrasonography can be used to determine whether an abscess exists, since diffuse inflammation will show an enlarged gland without a focal mass, whereas an abscesswill produce a discrete, fluid-filled, echo-free area that is well demarcated from adjacent glandular tissue38,39,43,44 Scattered internal echoes may also be visible, reflecting the presence of semisolid necrotic material within the abscesscavity.43,45If the presence of pus is established, ultrasonography can guide aspiration for the purpose of draining the abscessand obtaining material for culture.46 Furthermore, since the method is quick and easily reproducible, it can also be used to monitor responseto therapy. Pickrell and associates4’ were the first to use ultrasonography for preoperative localization of a parotid gland calculus. Recent studies indicate that ultrasonography is a fairly reliable method for demonstrating salivary calculi, since more than 90% of stones larger than 2 mm can be detected as echodense spots casting a characteristic acoustic shadow.39,48To some extent distinguishing between an intraglandular and an intraductal location is also possible with this procedure. Nevertheless, plain radiography and sialography remain the methods of choice for evaluating salivary calculi, because these methods are on the whole more accurate than ultrasonography for both detection and precise localization of calculi. The use of diagnostic ultrasound should therefore be restricted to selected cases, for

632

van den Akker

Oral November

Surg 198X

Fig. 10. Lipoma of left cheek. A, Well-defined mass with density characteristic of fat anterior to the masseter muscle (arrow). B and C, Sialography reveals anterolateral displacement of Stensen’s duct (arrows). Note opacified parotid gland (P).

instance patients with signs and symptoms of calculus obstruction in whom a salivary duct cannot be cannulated (Fig. 6). COMPUTED

TOMOGRAPHY

CT of the salivary glands was introduced as a new diagnostic technique in 1978,4gand subsequent experiences have firmly established the procedure as the method of choice for evaluating space-occupying lesions in and around the major salivary glands.50-57 CT findings in inflammatory and nonneoplastic diseaseshave also been reported,5*,5gbut most authors agree that the results are generally nonspecific and may sometimes be confused with malignant disease.54~60 Therefore, if radiologic examination is indicated in these cases,plain radiography and sialography remain the investigations of choice.5L.s2,55,60*6’ The major advantages of CT are its ability to directly reveal soft and hard tissues and to demonstrate small differences in soft tissue density. Axial and coronal planes can be used to visualize the parotid and submandibular glands, but in most cases the straight axial projection provides excellent anatomic detail and obviates the need for an additional coronal view. In parotid gland imaging, however, it may be necessary to slightly change the angle of the axial projection to avoid artifacts from dental restorations.56,57 On CT scans the normal parotid gland is readily discernible, being more radiodense than adjacent fat but less dense than surrounding muscles because of its high fat content and high concentration of saliva (Fig. 7). However, parotid gland density is variable; the parotid glands are less dense in older and heavier individuals whose glands are more infiltrated with fat. The submandibular gland contains less fat than the parotid gland so that its density is similar to the

density of adjacent muscles.57However, this gland is most easily identified by its characteristic shape and location (see Fig. 8). On plain CT images most benign tumors within the parotid gland appear as sharply marginated masses. As a rule these tumors are hyperdense in relationship to the surrounding parenchyma; hypodenselesions can also be observed but less frequently (Fig. 9). In contrast to benign tumors malignant lesions are reported to have poorly defined margins.5o,52They may also extend beyond the gland, obliterating adjacent fat and fascial planes.54*56 However, several authors have observed some malignant tumors to be well defined and marginated.54,61 Therefore differentiation between benign and malignant massesis not in fact always possible by means of CT, and tumor type cannot always be predicted.s3,62 Exceptions are fat-containing tumors such as lipomas and liposarcomas, which show on CT as densities characteristic of fat5’,53* 55,57(Fig. lo), and vascular lesions such as hemangiomas and chemodectomas, which strikingly enhance after intravenous infusion of contrast materia1.5’~55 The submandibular gland has a higher density than the parotid gland so that a submandibular gland tumor may have nearly the same density as the gland itself and therefore may be difficult to identify. However, comparison with the contralateral normal gland will generally show a difference in size or contour or both that indicates the presence of a space-occupying lesion (Fig. 8). Nevertheless a combination of CT and sialography may sometimes be necessary to clearly outline a submandibular gland tumor.56 Becauseof the slight density difference between a tumor and the surrounding normal salivary tissue, most early authors recommended a combination of CT and sialography rather than CT alone to improve

Volume 66 Number 5

Diagnostic imaging in salivary gland disease

633

Fig. 11. A, Axial MR scan (spin-echo image; repetition time = 700 msec; echo time = 30 msec) of a patient with an adenoid cystic carcinoma in the tail of the left parotid gland (white arrow]. The tumor is well separated from the internal carotid artery and internal jugular vein (black arrows~. B, Coronal MR scan (spin-echo image; repetition time = 600 msec; echo time = 30 msec) shows medial displacement of the internal jugular vein by the tumor (arrow).

localization of a mass and to better define its margins.49-53 Furthermore sialographic delineation of the glandular borders was considered essential to differentiate intrinsic from extrinsic lesions, particularly deep lobe parotid gland tumors from parapharyngeal masses.50*52*53 Others, however, have shown that with the newer generation, high-resolution CT scanners intraductal contrast is generally not necessary unless plain CT provides insufficient information or unless a specific anatomic structure, such as Stensen’s duct, must be visualized54-56(Fig. 10). If sialography is performed, water-soluble contrast medium is preferable, since oily contrast material is denser than necessary and may obscure anatomic detail or even create artifacts.56A combination of CT and intravenous infusion of contrast material has also been advocated, because most salivary gland tumors, and certainly a hemangioma, will enhance with contrast medium.57 However, this combined procedure is primarily indicated in the case of a tumor not located within a salivary gland, particularly a parapharyngeal mass, to distinguish between a vascular lesion, such as a chemodectoma, and other types of tumor that occur in this region.54s56 Also, in selected cases enhancement of the major regional vessels(i.e., the carotid arteries and the retromandibular and internal jugular veins) may be helpful by demonstrating their relationship to a lesion either within or outside of the parotid gland. Probably the most important indication for CT is the clinical situation in which it cannot be established whether a mass is intrinsic or extrinsic to the

gland. This use is particularly important for a deep-seated lesion, jammed between the mandible and the mastoid, whose origin may be either in the deep lobe of the parotid gland or in the parapharyngeal tissues.50,52-55 The surgical approaches to the two locations differ in that removal of a deep lobe parotid gland tumor requires a transparotid approach involving complete dissection and retraction of the facial nerve, whereas an extrinsic parapharyngeal mass requires a transcervical approach.50,55 Therefore CT findings in these instances may be extremely useful not only for treatment planning but also for preoperative patient counseling with regard to expected morbidity. The anatomic structure most related to postoperative morbidity after parotid gland surgery is the facial nerve, whose path through the glandular parenchyma arbitrarily divides the gland into a superficial and a deep part. The nerve itself is not visible on axial CT scans,but theoretically the course of the nerve through the gland can be inferred from its spatial relationship to visualized landmarks such as the styloid process and the retromandibular vein, both structures being directly medial to the facial nerve. Some authors claim that in this way the position of the nerve in relation to a tumor can be accurately determined in 75% or more of patients. s2*54, 55,60However, as noted by Wiesenfeld and colleagues,6’these landmarks are only visible on scans of the superior part of the parotid gland and, moreover, may be variable or distorted because of the tumor. Furthermore, the retromandibular vein

634

van den Akker

Oral Surg November 1988

cannot always be identified on plain CT scans, so that intravenous contrast-medium infusion or concomitant sialography may be necessary to visualize this vessel clearly. Nevertheless, it is apparent that CT may be of some help in determining whether a tumor lies medial or lateral to the facial nerve. However, expectations should not run too high because in many cases it will not be possible to predict the position of the nerve with any certainty.56 MAGNETIC

RESONANCE

IMAGING

MRI is an exciting new imaging technique based on the phenomenon of nuclear magnetic resonance (NMR). The method is capable of producing highquality tomographic images of certain parts of the body, including the salivary glands, without the use of ionizing radiation (Fig. 11). In brief, the technique involves placing the patient in a strong magnetic field so that nuclei of certain elements will align themselves along the axis of this field. The patient is then exposed to a second, perpendicularly oriented magnetic field in the radiofrequency region of the electromagnetic spectrum; this second magnetic field causes the nuclei in the subject to resonate about their equilibrium position. After the radiofrequency pulse is switched off the nuclei return to their original equilibrium position while emitting a radiofrequency wave, which can be detected as an NMR signal. A number of nuclei have properties that enable them to be measured by NMR, but it is the combined sensitivity and abundance of hydrogen that makes MRI possible. The procedure therefore essentially provides information on the water content of various tissues. Because the hydrogen nucleus contains a single proton, the term proton density is used to describe the distribution of resonating nuclei within the patient.63 The relaxation processof the nuclei after removal of the radiofrequency pulse is characterized by the time constants Tl and T2, and the choice and timing of the pulse sequencesdetermine the amount of Tl or T2 information in the signal. This information and the proton density of the examined tissue are the data from which an image can be constructed. The sensitivity of MRI is such that small differences in water content are sufficient to distinguish between normal and pathologic tissue, and in this respect the procedure is superior to CT. However, as far as diagnostic specificity is concerned, both techniques are somewhat disappointing because tissue characterization still remains very difficult.64 Recent studies indicate that MRI offers several

advantages over CT in the diagnosis of parotid gland masses.In a preliminary comparitive evaluation of 12 parotid gland masses,Schaefer and colleaguesh’ assessedfive clinically important imaging parameters, including conspicuousnessof the lesion, marginal appearance, internal architecture, regional extension, and artifact degradation. They concluded that for all these parameters MRI was equal or superior to CT. Similarly Lloyd and Phelps66found MRI to be superior to CT for demonstration of tumors of the parapharyngeal space. Particularly in this location, the greater soft tissue contrast resolution of MRI and the possibility of obtaining direct, three-dimensional projections without repositioning of the patient result in a better delineation of the tumor from surrounding soft tissue structures. Furthermore, flowing blood in MRI will not produce a signal, becausethe activated protons will have moved away from the imaging plane before their emitted signal is detected. The major vesselsin the parotid gland and the parapharyngeal space are therefore clearly recognizable without the use of contrast agents so that their relationship to the mass can easily be appreciated6” (Fig. 11). The absence of streak artifacts resulting from bone-related x-ray scatter and from dental restorations (amalgam and gold) is another important advantage of MRI over CT in the investigation of salivary gland disease.65.67However, ferromagnetic stainless steel in dentures and orthodontic braces will produce extensive artifacts in MRI, and if possible, these appliances should be removed before imaging.67 A disadvantage of MRI is the slow patient throughput resulting from the long sequenceacquisition time. Furthermore the procedure is contraindicated in patients with cardiac pacemakers or with ferromagnetic hemostatic clips in whom a risk of dislodgment with resulting hemorrhage or of injury of sensitive adjacent structures such as the brain exists.67,68 A recent study indicates that with stapedectomy prostheses there is no danger of displacement,69 but further research is required before patients with other types of metal implants can safely be exposed to MRI. DISCUSSION

For more than 50 years sialography and plain radiography were the only modalities available for salivary gland imaging. In the last decadesscintigraphy, ultrasound, CT, and, very recently, MRI have been introduced for this purpose, and there is no doubt that these procedures have considerably widened the diagnostic scope. However, the marked

Volume 66 Number 5

increase in imaging techniques has also introduced the problem of selecting the appropriate procedure for the individual patient. Furthermore, a danger of overuse exists, and it should be realized that not every clinical situation automatically requires an imaging procedure. CT scanning, for instance, is not necessary when a tumor has been diagnosed as a pleomorphic adenoma by cytologic studies and is clinically palpable in the superficial lobe of the parotid gland. The relative abundance of imaging techniques presently available makes a full appreciation of their diagnostic value absolutely essential for a wellconsidered choice. Sialography provides detailed information on ductal structure that is not otherwise obtainable. Sialography remains therefore the method of choice in cases of sialoadenitis, sialosis, and obstruction but, as a sole procedure, has become obsolete in tumor diagnosis since the advent of CT scanning. Scintigraphy with yymTc-pertechnetateis primarily a function test and has the advantage that all major salivary glands can be examined simultaneously and continuously over a period of time. The method is particularly useful in systemic diseases involving the salivary glands, such as SjBgren’s syndrome, because it objectively demonstrates the full extent of glandular dysfunction. Its value in tumor diagnosis and in infections and obstructive disorders is limited, but nevertheless the procedure may provide significant information in selected cases, A special indication for scintigraphy exists when the ductal orifice of a major salivary gland cannot be found or cannot be cannulated. This problem may occur in a variety of clinical conditions and makes scintigraphy the only diagnostic procedure available to confirm the diagnosis and to assess the remaining glandular function. The potential of ultrasonography for detecting salivary gland disease often appears to be overlooked. In certain clinical situations the method is competitive with CT, and considering its advantages, ultrasonography should be used more often as an initial screening procedure, for instance to distinguish between a diffuse inffammatory process and a mass lesion or between a cyst and a solid tumor. Furthermore, ultrasound is a procedure that accurately defines the presence and extent of abscess formation in acute suppurative sialoadenitis. The principal advantage of CT is the ability to directly reveal the salivary glands and neighboring structures so that the location and extent of mass lesions can be accurately demonstrated. For this reason the procedure is particularly useful to establish whether a tumor is intrinsic or

Diagnostic imaging in salivary gland disease

635

extrinsic to a gland. It should be remembered, however, that CT scanning is not capable of providing a specific diagnosis except for typical casessuch as lipoma or cyst and that in most instances the method does not reliably distinguish between benign and malignant masses. The newest technique is MRI, which accurately images the major salivary glands and adjacent soft tissues without the use of ionizing radiation. The main indication for MRI is evaluation of masslesions, particularly tumors of the parapharyngeal space, and preliminary studies indicate that in this respect the procedure is superior to CT scanning. Advantages of MRI over CT include better contrast resolution, absenceof artifact degradation from dental restorations, visualization of major vessels without intravenous injection of contrast material, and direct, three-plane imaging without patient repositioning. These advantages result in a better delineation of the tumor, and it appears therefore that MRI may very well replace CT for salivary gland imaging in the near future. In conclusion, the introduction of new techniques for salivary gland imaging has had a strong impact on the role of older procedures, whose use has considerably diminished, and has broadened the diagnostic possibilities, particularly in tumor diagnosis. However, as before, the diagnostic work-up in each patient should be individually tailored, a procedure that requires a critical attitude toward the value of the various methods in each particular case. REFERENCES

5.

6.

I. 8. 9. 10.

11. 12.

Holt JF. Sialography. Radiology 1957;68:584-5. Gates GA. Sialography and scanning of the salivary glands. Otolaryngol Clin North Am 1977;10:379-90. Adam EJ, Willson SA, Corcoran MO, Hobsley M. The value of parotid sialography. Br J Surg 1983;70: 108- 10. Stacey-Clear A, Evans R, Kissin MW, Kark AE, Wilkins R. Sialography does not alter the management of parotid spaceoccupying lesions. Clin Radio1 1985;36:389-90. Calcaterra TC, Hemenway WG, Hansen GC. Hanafee WN. The value of sialography in the diagnosis of parotid tumors. Arch Otolaryngol 1977;103:727-9. Lindvall AM. Radiographic and scintigraphic examinations in diseases of the major salivary glands. Dentomaxillofac Radio1 198O;(suppl 1):1-l 13. Yune HY, Klatte EC. Current status of sialography. AJR 1972;115:420-8. Snow GB. Tumours of the parotid gland. Clin Otolaryngol 1979;4:457-68. Schulz HG. Das Riintgenbild der Kopfspeicheldriisen. Leipzig: Johann Ambrosius Barth, 1969:40- 107. HIusler RJ. N’Guyen VT, Ritschard J, Montandon PB. Differential diagnosis of xerostomia by quantitative salivary gland scintigraphy. Ann Otol Rhino1 Laryngol 1977;86:33341. Rauch S. Die Speicheldrtisen des Menschen. Stuttgart: Georg Thieme Verlag, 1959:434-41. Langlais RP, Kasle MJ. Sialolithiasis: the radiolucent ones. ORAL SURG ORAL MED ORAL PATHOL 1975;40:686-90.

636

van den Akker

Oral Surg November 1988

13. Suleiman SI, Hobsley M. Radiological appearances of parotid duct calculi. Br J Surg 1980;67:879-80. 14. Seward GR. Anatomic surgery for salivary calculi. II. Calculi in the anterior part of the submandibular duct. ORAL SURG ORAL MED ORAL PATHOL 1968;25:287-93.

15. Seward CR. Anatomic surgery for salivary calculi. IV. Calculi in the intraglandular part of the submandibular duct. ORAL SUKG ORAL MED ORAL PATHOL 1968;25:670-8.

16. Van den Akker HP, Busemann-Sokole E. Submandibular gland function following transoral sialolithectomy. ORAL SURG ORAL MED ORAL PATHOL 1983;56:35

l-6.

17. O’Riordan B. Phleboliths and salivary calculi. Br J Oral Surg 1974;12:1 19-31. 18. Van der Waal I, Beemster G, Kraayenhagen HA, van der Kwast WAM. Pilomatrixoma in the differential diagnosis of pathologic calcification of the paraoral soft tissue. Int J Oral Surg 1977;6:283-6. 19. Markitziu A, Sela M, Seltzer R. Major salivary glands in branchial arch syndromes. ORAL SURG ORAL MED ORAL PATHOL I984;58:672-7.

20. Van den Akker HP, Busemann-Sokole E. Absolute indications for salivary gland scintigraphy with 99mTc-pertechnetate. ORAL SURG ORAL MED ORAL PATHOL 1985;60:440-7.

21. Busemann-Sokole E, van den Akker HP, van der Schoot JB. Sequential salivary gland scintigraphy using technetium-99m and carbachol: a clinical test. In: Medical radionuclide imaging, vol 2. Vienna: International Atomic Energy Agency, 19771363-9. 22. Parret J, Peyrin JO. Radioisotopic investigations in salivary pathology. Clin Nucl Med 1979;4:250-61. 23. Schall CL. The role of radionuclide scanning in the evaluation of neoplasms of the salivary glands: a review. J Surg Oncol 1971;3:699-714. 24. Magnus L, Schmitt Cl, Striitges MW, Lehmann Cl. Die Tumordiagnostik der Ohrspeicheldrtisen mittels Sialographie und Sialosrintigraphie. Strahlentherapie 1979;155:171-4. 25. Mishkin FS. Radionuclide salivary gland imaging. Semin Nucl Med 1981;11:258-65. dose-response analysis of salivary 26. Tsujii H. Quantitative function following radiotherapy using sequential RI-sialography. Int J Radiat Oncol Biol-Phys 1985-l 1:1603-12. 27. Newman RK. Weiland FL. Johnson JT. Rosen PR. Gumerman LW. Salivary scan after major ablative head and neck surgery with prediction of postoperative fistulization. Ann Otol Rhino1 Laryngol 1983:92:366-g. 28. Schall GL, DiChiro G. Clinical usefulness of salivary gland scanning. Semin Nucl Med 1972;2:270-7. 29. Greyson ND, Noyek AM. Radionuclide salivary scanning. J Otolaryngol 1982; I 1(suppl 10): l-47. 30. Bornemann C, Creutzig H. Die klinische Bedeutung einer Speicheldriisenszintigraphieeine prospektive Studie. HNO 1983;3 1:200-6. 31. Noodt A, Schneider P, Traurig G, Draf W, Haas JP. Quantitative Funktionsszintigraphie der Speicheldriisenklinische Moglichkeiten und derzeitige Indikationen. HNO 1985:33:232-g. 32. Schall GL, Anderson LG. Wolf RO, et al. Xerostomia in Sjogren’s syndrome-evaluation by sequential salivary scintigraphy. JAMA 1971;216:2109-16. 33. Alarcon-Segovia D, Ibanez G, Hernandez-Ortiz J, Cetina JA, Gonzalez-Jimenez Y, Diaz-Jouanen E. Salivary gland involvement in diseases associated with SjBgren’s syndrome. 1. Radionuclide and roentgenographic studies. J Rheumatol 1974;1:159-65. 34 Daniels TE, Powell MR. Sylvester RA, Talal N. An evaluation of salivary scintigraphy in Sjogren’s syndrome. Arthritis Rheum 1979;22:809-14. 35. Yamashita T. Chiyonori I, Tomoda K, Kumazawa T. Prognostic determination and submandibular function in Bell’s palsy--dynamic testing with technetium-99m. Arch Otolaryngol 1985:111:244-8.

36. Kaneko T, Kobayashi N, Miura T, Asano H, Kitamura T. L’echographie ultrasonique pour l’exploration des tumeurs parotidiennes. Ann Otolaryngol Chir Cervicofac 1975; 92:685-90. 37. Neiman HL, Phillips JF, Jaques TL, Brown TL. Ultrasound of the parotid gland. J Clin Ultrasound 1976:4:11-3. 38. Rothbdrg R, NGyek AM, Goldfinger M, Kassel EE. Diagnostic ultrasound imaging of parotid disease-a contemporary clinical perspective. J Otolaryngol 1984;13:232-40. 39. Wittich GR, Scheible WF, Hajek PC. Ultrasonography of the salivary glands. Radio] Clin North Am 1985;23:29-37. 40. Bruneton JN, Sicart M, Roux P, Pastaud P, Nicolau A, Delorme G. Indications for ultrasonography in parotid pathologies. ROFO 1983;138:22-4. 41. Ballerini G, Mantero M, Sbrocca M. Ultrasonic patterns of parotid masses. J Clin Ultrasound 1984;12:273-7. 42. Caste1 JC, Delorme G. Semiologie echographique des tumeurs de la parotide et correlations histo-ichographiquesa propos de soixante-dix cas. Ann Radio] (Paris) 1984;28:3604. 43. Pirschel J. Die Erkrankungen der Parotis im hochauflosenden real-time Schnittbild. ROFO 1982;137:503-8. 44. Wagner A, Schadel A, Bottcher HD. Die Diagnostik raumfordernder Prozesse in der Glandula parotis durch Ultraschall. Dtsch Z Mund Kiefer Gesichtschir 1985;9:55-60. 45. McCurdy JA, Nadalo LA, Yim DWS. Evaluation of extrathyroid masses of the head and neck with gray scale ultrasound. Arch Otolaryngol 1980;106:83-7. 46. Magaram D, Gooding GAW. Ultrasonic guided aspiration of parotid abscess. Arch Otolaryngol 1981;107:549. 47. Pickrell KL, Trought WS, Shearin JC. The use of ultrasound to localize calculi within the parotid gland. Ann Plast Surg 1978;1:542-6. 48. Gritzmann N, Hajek P, Karnel F, Fezoulidis J, Turk R. Sonographie bei Speichelsteinen-Indikation und Stellenwert. ROFO 1985;142:559-62. 49. Carter BL, Karmody CS. Computed tomography of the face and neck. Semin Roentgen01 1978;13:257-66. 50. Som PM, Biller HF. The combined CT-sialogram. Radiology 1980;135:387-90. 51. Carter BL, Karmody CS, Blickman JR. Panders AK. Computed tomography and sialography. 2. Pathology. J Comput Assist Tomogr 1981;5:46-53. 52. Stone DN, Mancuso AA, Rice D, Hanafee WN. Parotid CT sialography. Radiology 1981;138:393-7. 53. Schindler E, Reck R. Die Kombination von Computertomographie und Sialographie zur Parotisdiagnostik. Radiologe 1982;22:241-6. 54. Bryan RN, Miller RH, Ferreyro RI, Sessions RB. Computed tomography of the major salivary glands. AJR 1982;139:54754. 55. McGahan JP, Walter JP, Bernstein L. Evaluation of the parotid gland--comparison of sialography, non-contrast computed tomography, and CT sialography. Radiology 1984; 152:453-g. 56. Rabinov K, Kell T, Gordon PH. CT of the salivary glands. Radio] Clin North Am 1984;22:145-59. 57. Saluk PH, Swartz JD, Korsvik H, Marlowe FI. Highresolution computed tomography of the major salivary glands: current status. J Comput Assist Tomogr 1985;9:39-50. of the parotid gland: computed 58. Zeit RM. Hypertrophy tomographic findings. AJR 1981;136:199-200. 59. Iko BO. Computed tomography and sialography of chronic pyogenic parotitis. Br J Radio1 1984;57:1083-90. 60. Eyjolfsson 0, Nordshus T, Dahl T. Sialography and CTsialography in the diagnosis of parotid masses. Acta Radial (Diagn) 1984;25:361-4. 61. Wiesenfeld D, Ferguson MM, McMillan NC. Simultaneous computed tomography and sialography of the parotid and submandibular glands. Br J Oral Surg 1983;21:268-76. 62. Mooyaart EL, Panders AK, Vermeij A. CT scanning of

Volume 66 Number 5

63. 64. 65. 66. 67.

tumors in or near the parotid or submandibular glands. Diagn Imaging Clin Med 1984;53:177-S1. Smith MA. The technology of magnetic resonance imaging. Clin Radio1 1985;36:553-9. Steiner RE. Magnetic resonance imaging: its impact on diagnostic radiology. AJR 1985;145:883-93. Schaefer SD, Maravilla KR, Close LG, Burns DK, Merkel MA, Suss RA. Evaluation of NMR versus CT for parotid masses:a preliminary report. Laryngoscope 1985;95:945-50. Lloyd GAS, Phelps PD. The demonstration of tumours of the parapharyngeal space by magnetic resonance imaging. Br J Radio1 1986;59:675-83. New PFJ, Rosen BR, Brady TJ, et al. Potential hazards and artifacts of ferromagnetic and nonferromagnetic surgical and dental materials and devices in nuclear magnetic resonance imaging. Radiology 1983;147:139-48.

Diagnostic imaging in salivary gland disease

637

68. Kean DM, Worthington BS, Firth JL, Hawkes RC. The effects of magnetic resonance imaging on different types of microsurgical clips. J Neural Neurosurg Psychiatry 1985; 481286-7. 69. Applebaum EL, Valvassori GE. Effects of magnetic resonance imaging fields on stapedectomy prostheses.Arch Otolaryngol 1985;111:820-1. Peprint requesfs to:

Dr. H. P. van den Akker Department of Oral and Maxillofacial Surgery Academic Medical Center Meibergdreef 9 1105 AZ Amsterdam The Netherlands