Ultrasound is the first and basic imaging method employed in the case of diseases affecting superficial tissues and organs, including the salivary glands. The article discusses basic mistakes that can be made during an ultrasound examination of these structures.
The following reasons may account for mistakes in ultrasound imaging of the salivary glands:
difficult conditions during examination (technical or patient-related);
the lack of knowledge of normal anatomy (doctor-related);
the lack of familiarity with characteristics of ultrasound images in various salivary gland diseases (doctor-related);
the similarity of ultrasound images in different diseases;
the lack of knowledge of statistical incidence of diseases, but also attaching excessive importance to such statistical data (doctor-related);
the lack of clinical and laboratory data as well as the lack of results of other examinations, their insufficient number or incorrectness.
Difficult conditions of examination
An ultrasound examination may be significantly hindered by short and thick neck of the examined patient, poor tissue transparency for ultrasounds or limited neck mobility, making it impossible to tilt the head as far back as needed or turn it sideways. What greatly impedes and at times even prevents a reliable examination is the lack of contact with the patient or cooperation on their part, e.g. patient's moving during an examination. The clinical condition of the examined patient, e.g. rapid, deep breathing, may impede or preclude a reliable Doppler evaluation.
Tissue transparency for ultrasounds is worse in some patients than in others. This can be due to, among other things, obesity, postoperative changes or other reasons for fibrosis and tissue architectonics dysfunctions. Subcutaneous emphysema can make it impossible to examine tissues located deeper.
The description of the examination should include information about the presence and nature of encountered difficulties to make the referring doctor aware of potential limitations of the examination reliability and, in the case of incompatibility with the clinical picture, to make it possible for them to decide on the need for additional examinations, such as computed tomography or magnetic resonance imaging.
Mistakes resulting from the lack of knowledge of normal anatomy
The key element necessary for reliable ultrasound evaluation of the area of the salivary glands and lymph nodes, and basically the entire area of the neck and face, is the knowledge of normal anatomy. A large number of anatomical structures in this area, their course and arrangement as well as complicated mutual spatial relations make an ultrasound image of the area complex, and the lack of knowledge about these issues precludes proper performance of an examination and interpretation of an image. In addition, an ultrasound image of the discussed areas may be changeable, e.g. may depend on the history of neck irradiation or content of fat tissue. One of correct anatomical structures may be easily mistaken for a focal change. For instance, one of the muscles of the neck base, the genioglossus muscle in the transverse section in the area of the attachment to the bone of the mandible, may mimic a focal lesion that might even seem like a malignant one on ultrasound (Fig. 1). Another example: anterior belly of the digastric muscle in the transverse or oblique section has an oval shape and may mimic a well-delineated focal change (Fig. 1: structure No 4). Proper performance of an examination, i.e. in at least two perpendicular planes, along with knowledge of the course and image of muscles in an ultrasound examination enables easy identification of the viewed structures.
The structure marked with arrows in the first image might be described as a heterogeneous focal lesion with blurred, irregular contours and outgrowths, i.e. with standard characteristics of a lesion that appears as malignant on ultrasound. In fact, it is the genioglossus muscle in the transverse section (1). Correct ultrasound anatomy of the floor of the mouth area in the transverse section: (2) – geniohyoid muscle; (3) – mylohyoid muscle; (4) – anterior belly of the digastric muscle; (5) – tongue; (6) – sublingual gland; (7) – oral cavity; (8) – subcutaneous tissue
Yet another example of a misdiagnosis based on wrong interpretation of an anatomical ultrasound image is suggesting sialolithiasis, for instance on the basis of an image of fibrosis (e.g. transverse or oblique section through an artery wall), fragments of correct anatomical bone structures or air bubbles in the region of the floor of the oral cavity (Fig. 2, 3, 4, 5).
In the submandibular gland parenchyma (outlined with a dotted line), one can observe a hyperechoic linear structure (arrowheads) with acoustic shadow (arrows), i.e. a standard image of a deposit on a presented still image in the grey scale. During a dynamic examination, it can be seen that the hyperechoic line is an outer contour of a tubular structure that passes through the salivary gland in spirals and runs on upwards on the cheek and, below, toward the external carotid artery. The complete image suggests that it is a fragment of the facial artery wall affected by fibrosis, whereas the “acoustic shadow” is the vascular lumen captured in the oblique section. A Doppler examination, e.g. with the power Doppler, may serve as additional, unambiguous confirmation that the structure is the facial artery (Fig. 3)
Unambiguous confirmation with the power Doppler that the structure in Fig. 2 is the facial artery (arrowheads). The parenchyma of the submandibular gland is marked with arrows
Deep in the parenchyma of the parotid gland (2), a linear hyperechoic structure (arrows) can be seen which might be confused with a deposit. In fact, it is a fragment of the head of the mandible. The continuation of an echo of the anterior contour of the mandible ramus is marked with arrowheads; 1 – subcutaneous tissue
The hyperechoic linear structure (arrows) in the area of the mouth floor is not a deposit in the excretory duct of the gland (Wharton's duct) but rather an echo at the boundary of air bubbles mixed with saliva. Arrowheads – mandibular gland
The following may present themselves as a deposit (stone) on an ultrasound image(1):
gas (e.g. air bubbles);
Small salivary glands
Small glands are located in the oral mucosa and submucosa of the mouth, palate, nasal sinuses, throat, larynx, trachea and bronchi as well as in the tongue area. Tumors arising from the parenchyma of the salivary glands can also be located there and should be taken into account in differential diagnosis(2). In the case of small salivary glands, malignant tumors definitely prevail.
Deep lobe of the parotid gland and fatty infiltration of the salivary glands
In the case of the parotid gland, it is impossible to visualize and thus evaluate potential focal changes in the deep lobe. Penetration of ultrasounds into the parotid gland and even into deeper areas of the submandibular gland is further hindered if there is high content of fat in the parenchyma. Attenuation of ultrasounds is then significant and echogenicity of the parenchyma is raised and blurred (Fig. 6). If the entire focal change cannot be seen in an ultrasound examination or if a malignant lesion is suspected, magnetic resonance imaging or computed tomography should be ordered so as to visualize the entire lesion or evaluate lymph nodes that are located deeper or are inaccessible for ultrasound.
The first image shows the submandibular gland (arrows) with low fat content and sharp borders. Organs and tissues located in the vicinity and below the submandibular gland can be clearly seen. (1) – palatine tonsil; (2) – tongue; (3) – posterior belly of the digastric muscle; (4) – mylohyoid muscle; (5) – hyoglossus. The second image presents the submandibular gland (arrowheads) with high fat content – the anterior contour of the parenchyma is hardly visible, whereas the posterior contour of the gland cannot be unambiguously traced. Structures located deeper are not visible
The borders of the salivary glands, in particular the parotid glands but also the submandibular ones, may be difficult to trace even if their condition is correct, especially when steatosis of the parenchyma occurs, which further blurs echogenicity in the center and contours of the glands causing segmental „merging” with the surrounding connective tissue, e.g. the subcutaneous one. For this reason, it may be difficult to unambiguously locate focal changes in the parenchyma of the salivary gland or outside of it, particularly in the region of the poles of the parotid glands and within the submandibular area (Fig. 7, 8, 9). One should also pay attention to lesions in neighboring organs so as not to describe their pathology as a salivary gland disease, e.g. changes in the mandible, muscles, tonsils or subcutaneous tissue(3, 4).
Focal lesion (arrows) in the area of the lower pole of the parotid gland (SP) on the panoramic ultrasound image. The location of the lesion is difficult to determine unambiguously. The fact that the sternocleidomastoid muscle (M) penetrates above the lesion suggests that it is likely that the lesion is located outside of the parenchyma of the gland. Branchial cleft cyst (following FNAB and postoperative histopathological examination)
Focal lesion (arrows) in the area of the lower pole of the parotid gland (SP) on the panoramic ultrasound image. The location of the lesion is difficult to determine unambiguously. However, it seems that it is surrounded by the parenchyma of the gland from the above. M – sternocleidomastoid muscle; T – common carotid artery. Wharton's tumor (following FNAB and postoperative histopathological examination)
Reactive lymph node (arrows) with central blood flow (regular, symmetrical vessel segments – ramification of the longitudinal vessel) in the submandibular area (2), above the submandibular gland. Yet, it is difficult to unambiguously trace the outer contour of the submandibular gland; 1 – subcutaneous tissue
Intrasalivary lymph nodes
There may be lymph nodes located in the parenchyma of the salivary glands. They should not be described as pathology if they present correct ultrasound features(5).
Small focal changes
Another significant issue is differentiating small (not exceeding 20 mm in diameter) focal changes of the salivary glands(6, 7). Even if they are oval, well-delineated and have regular contours as well as a uniform structure, i.e. even if they present a standard ultrasound image of a benign lesion, their malignant nature cannot be ruled out. Small malignant neoplastic changes of the salivary glands and metastases to salivary glands may present an image of a benign lesion in an ultrasound examination (Fig. 10). For this reason, the safest approach in the case of focal changes in the salivary glands is their verification by an ultrasound-guided fine-needle aspiration biopsy (FNAB).
Oval, quite well-delineated (high-resolution probe) focal lesion (arrows) with posterior acoustic enhancement. Mucoepidermoid carcinoma in an ultrasound-guided fine-needle aspiration biopsy
Certain symptoms or clinical data, such as pain, which may suggest infiltration of nerves, or the patient's age, may further contribute to the urgent need for a biopsy. The necessity for this procedure is of course undisputed if the ultrasound image itself suggests a malignant lesion.
Changes in the entire parenchyma of the salivary gland
A large hypoechoic focal change occupying the entire or almost the entire parenchyma of the salivary gland may be mistaken for inflammation (!) due to its statistically higher incidence. Even focal changes of considerable sizes may have regular margins without any outgrowths, which does not rule out their malignant nature. They can infiltrate neighboring tissues and structures, e.g. bones, and therefore one should always pay attention to whether the contours of bone surfaces seen in an ultrasound image are regular and correct, according to their anatomical shape (Fig. 11).
Hypoechoic, well-delineated focal lesion (arrowheads) causing clear destruction of the mandible ramus (arrows); only a narrow streak of the preserved correct parenchyma of the parotid gland can be seen around the lesion (anterior contour of the superficial lobe marked with hollow arrows). Acinic-cell carcinoma
Invasive neoplasms destroying the entire parenchyma of the salivary gland may have blurred borders but in the case of chronic inflammation, contours of the salivary glands can also be irregular (Fig. 12). If irregular finger-like outer contours of the salivary gland have been visualized, ultrasound-guided FNAB is the procedure to be followed.
Changes in the view of the submandibular glands in two ultrasound images (power Doppler examination) of two different patients look similar: uneven margins, blurred contours, irregular shape, heterogeneous echogenicity, single flow segments. The inflamed parenchyma of the submandibular gland in the first US image (arrows) and adenoid cystic carcinoma whose infiltration destroyed the entire parenchyma of the salivary gland in the second US image (arrowheads)
One should always look for the correct parenchyma of the salivary gland around a large focal change located close to the salivary glands as this procedure makes it easier to correctly suggest a focal lesion rather than inflammation (Fig. 11). There are two reasons why the visualization and identification of a correct fragment of the parenchyma can be troublesome: the above-mentioned difficulty in tracing contours of the salivary glands, even healthy ones, as well as a small size of the preserved fragment of the parenchyma. One should also search for signs of infiltration of the neighboring structures, e.g. bones or muscles as well as lymph nodes having an abnormal ultrasound image, which could suggest a malignant nature of the lesion.
What should be alarming to the examining doctor is the lack of clinical symptoms of sialolithiasis and inflammation as well as the lack of deposits seen during an ultrasound when hypoechoic and heterogeneous salivary glands have been found. This particularly applies for the submandibular glands because the incidence of malignant neoplasms in this area is close to 50%(1). However, the co-occurrence of sialolithiasis does not automatically rule out a neoplasm of the salivary gland. On the other hand, a focal change in the sublingual gland is in the vast majority of cases a malignant neoplasm. In case of any doubts, it is best to perform FNAB.
Pseudocyst – a misdiagnosis of a cyst
One should exercise particular caution when suggesting the presence of a simple cyst in the head and neck area since many other pathologic changes, including malignant neoplasms, may look very much the same on ultrasound. Oval, well-delineated shape, anechoic center and clear posterior acoustic enhancement, i.e. standard characteristics of a simple cyst on ultrasound, may be very misleading in organs and soft tissues of the face and neck(1, 8–11). Primary reasons for that include: statistically low incidence of simple cysts in the salivary glands and neck, a number of changes that may mimic the image of a simple cyst and the fact that, in the considerable majority of cases, branchial cleft cysts paradoxically present themselves as solid lesions on grey-scale ultrasound.
Each of the following may present standard features of a simple cyst(1, 8–11):
lymph nodes in lymphomas;
benign and malignant neoplasms of the salivary glands;
metastatic lymph nodes;
reactive lymph nodes;
benign lymphoepithelial lesions in HIV-positive patients.
The lower the resolution of the ultrasound probe and the lower the grey scale available in the ultrasound machine used by the examining doctor, the more suggestive the imitation of a cyst by other histopathological focal changes on ultrasound.
Not only may benign and malignant neoplasms of the salivary glands mimic a cystic lesion when their interior is filled with fluid and the layer of cancer cells is located only on the periphery, but also when they are completely solid. In some cases, the use of sensitive color Doppler or power Doppler imaging may unequivocally suggest a proper diagnosis. Still, neoplasms of the salivary glands can be poorly vascularized and even with the use of sensitive power Doppler, the flow inside a solid lesion might not be seen (Fig. 13).
Oval, well-delineated, nearly anechoic focal lesion (arrowheads) with posterior acoustic enhancement in the parotid gland (SP). Color Doppler showed no evidence of flow (the option is not visible in the picture). The image may suggest a simple cyst, when in fact it is pleomorphic adenoma. Arrows – anterior contours of the mandible; T – external carotid artery; Z – retromandibular vein
In the case of lesions that are indeed filled with fluid whose nature is determined solely by the type of lining cells in the peripheral part, even the highest-class ultrasound machine may not be helpful in making a proper diagnosis. The preoperative examination that may enable their verification is FNAB.
Lymph nodes that may misleadingly appear on an ultrasound as a simple cyst may occur in the typical position in the neck (e.g. in the submandibular area, along the edges and underneath the sternocleidomastoid muscle, along the large neck vessels), but also in the area of the parotid gland parenchyma(1).
Neoplastic infiltration vs inflammatory infiltration
Both inflammatory infiltrations and neoplastic infiltrations of the salivary glands and soft tissues may have a very similar image on ultrasound. An irregularly shaped, heterogeneous hypoechoic area with outgrowths may be typical of both types of infiltration. As neoplastic infiltration, inflammatory infiltration may spread onto neighboring tissues, e.g. the sternocleidomastoid muscle.
A certain type of salivary gland inflammation, the so-called Küttner's tumor, may mimic a malignant lesion in a clinical examination – it is hard. Similarly, it can present blurred borders and heterogeneous structure on ultrasound(12, 13).
The image of oval, hypoechoic or nearly anechoic lesions disseminated in the parenchyma may occur in the case of(8, 14–17):
Sjögren's syndrome (Fig. 14);
granulomatous disease (e.g. tuberculosis, sarcoidosis);
benign lymphoepithelial lesions in HIV-positive patients.
A. Multiple, nearly anechoic, oval focal lesions disseminated in the parenchyma of the parotid gland (symmetrical in both glands). B. Power Doppler shows increased blood flow. Sjögren's syndrome. The image is not pathognomonic
Multiple solid lesions
The image of solid lesions disseminated in the parenchyma may be caused by(4, 18–20):
Warthin's tumor (lymphatic cystadenoma);
recurrent pleomorphic adenoma (tumor mixtus);
acinic cell carcinoma;
correct or affected lymph nodes.
Ultrasonography is the first imaging method applied in the case of diseases of the salivary glands. The article discussed basic mistakes that can be made during an ultrasound examination of these structures. The reasons for these mistakes may be examiner-dependent or may be beyond their control. The latter may include, inter alia, difficult conditions during examination (technical or patient-related), similarity of ultrasound images in different diseases, the lack of clinical and laboratory data as well as the lack of results of other examinations, their insufficient number or incorrectness. Doctor-related mistakes include: the lack of knowledge of normal anatomy, characteristics of ultrasound images in various salivary gland diseases and statistical incidence of diseases, but also attaching excessive importance to such statistical data.
The complex anatomical structures of the floor of the oral cavity may be mistaken for benign or malignant tumors. Fragments of correct anatomical structures (bones, arterial wall fibrosis, air bubbles in the mouth) can be wrongly interpreted as deposits in the salivary gland or in its excretory duct. Correct lymph nodes in the parotid glands may be treated as pathologic structures. Lesions not being a simple cyst, e.g. lymphoma, benign or malignant tumors of salivary glands, metastatic lymph nodes, can be mistaken for one. The image of disseminated focal changes, both anechoic and solid, is not pathognomonic for specific diseases in the salivary glands. However, in part, it occurs typically and requires an extended differential diagnosis. Small focal changes and infiltrative lesions pose a diagnostic problem because their etiology cannot be safely suggested on the basis of an ultrasound examination itself. The safest approach is to refer patients with abnormal focal changes for an ultrasound-guided fine-needle aspiration biopsy.
In conclusion, it should be stated that diagnostic pitfalls in an ultrasound examination can happen to anyone, even if the highest-class machine is used. However, one should aim at decreasing the number and, ultimately, at total elimination of mistakes resulting from the lack of knowledge. According to the authors, the necessity to continuously obtain new information applies to everyone, regardless of the level of knowledge that they already possess.
Conflict of interest
Authors do not report any financial or personal connections with other persons or organizations, which might negatively affect the contents of this publication and/or claim authorship rights to this publication.
This paper was prepared based on Ewa J. Białek, “Błędy i pomyłki w diagnostyce USG ślinianek i węzłów chłonnych położonych powierzchownie”, In: „Błędy i pomyłki w diagnostyce ultrasonograficznej”, ed. by Wiesław Jakubowski, Roztoczańska Szkoła Ultrasonografii, Warszawa – Zamość 2005: p. 45–58.