Keywords

1 Introduction

Ultrasonography has become widely utilized as a point-of-care diagnostic tool in clinical practice because of its noninvasive and nonirradiating nature, affordable equipment, simple maneuver, high-resolution image, and rescanning at ease in any setting. Anatomical characteristics of major salivary glands (SGs) are well suited to ultrasonography since parotid glands (PGs) and submandibular glands (SMGs) are superficially located with unique echogenicity and vascularity. Understanding normal and abnormal echostructures of major SGs is an essential requirement for salivary gland ultrasonography (SGUS).

The first report of SGUS appeared in the late 1960s [1]. A homogenous pattern and high reflectivity related to multiple, evenly distributed fibrous septa in the normal SGs were described in 1970 [2]. In the 1980s, a B-mode scanning technique described a heterogeneous finding with tiny liquid-like structures, volumetric atrophy, and decreased echogenicity in primary Sjögren’s syndrome (pSS) [3, 4]. Homogeneity distinguished SGUS findings of normal subjects from the adults with pSS reported in the 1990s [5]. Similar sonographic changes in juvenile Sjögren’s syndrome (jSS) in pediatric patients were reported [6, 7].

Several semiquantitative scoring systems for diagnosing pSS were developed in the past two decades. Sublingual glands (SLGs), the smallest salivary gland of the three pairs of major SGs, were not included in any scoring system since SLGs are too small and varied in size for a reliable evaluation. The scoring systems’ variation with different pathological definitions and inconsistency in interpreting SGUS findings yielded diverse diagnostic accuracy [8, 9]. Recently, normal and abnormal PG and SMG features in pSS were defined and validated [10]. A new standardized semiquantitative score was also developed [11] to overcome the concerns of the variation. Implementing SGUS scoring systems as one of the 2016 ACR/EULAR classification criteria for pSS increased the criteria’s sensitivity [12,13,14].

2 Normal Salivary Gland Sonographic Anatomy

2.1 Parotid Glands (PGs)

PGs are the largest SGs located in the parotid space. The parotid space formed by the superficial layer of the deep cervical fascia occupies the space lying below the external auditory canal and the mastoid process’s apex and above the mandible angle. The parotid space is lateral to the parapharyngeal space and the carotid space, medial to the superficial space and the subcutaneous tissue, anterior to the carotid space, superior to the submandibular space, and posterior to the masticator space [15]. The PG fills approximately two-thirds of the parotid space. Other structures in the space are Stenson’s duct, intraparotid lymph nodes, external carotid artery, retromandibular vein, facial nerve, and auriculotemporal branches of the mandibular division of the trigeminal nerve. Three structures, including the external carotid artery, retromandibular vein, and facial nerve, run parallel, in which the vein is in the middle.

The PG is anatomically divided by the intraparotid facial nerve into two-thirds or superficial lobe laterally to the nerve and one-third or deep lobe medially. The superficial lobe is the portion that overlies the preauricular region, ramus of the mandible, masseter muscle, and posterior mandibular surface. The gland extends superiorly to the zygomatic arch, inferiorly to the sternocleidomastoid muscle (SCM), and posteriorly to the superior border of SCM and mastoid tip (Fig. 10.1). The deep lobe is a small portion behind and deep to the mandibular ramus and lies within the parapharyngeal space [16]. The average gland dimensions are 5.8 and 3.4 cm in the craniocaudal and ventrodorsal axes [17]. The size of the gland is smaller in children and women. There is a high density of periglandular and intraglandular lymph nodes. The intraglandular lymph nodes are unique because of the parotid’s late encapsulation during embryogenesis. Approximately 90% of the nodes are periglandular, located between the glandular tissue and capsule [16]. Most of the intraglandular lymph nodes are mainly situated within the upper and lower poles of the glands.

Fig. 10.1
figure 1

Scanning of the normal parotid gland of a pediatric patient in the longitudinal (a, b, e) and transverse (c, d, f) planes. Children have smaller and slightly darker (less echogenic) gland than adults

Full size image

The normal PG parenchyma appears uniformly echotexture (Fig. 10.1b, d) comparable to normal thyroid parenchyma with a clear demarcation between the gland and the overlying tissue [11]. The parotid appears more echogenic than the surrounding muscles but less echogenic than the mandible ramus or mastoid process. In general, the echogenicity of the parenchyma depends on the amount of intraglandular fatty tissues. The gland in children appears smaller and less echogenic (darker) (Fig. 10.1e, f) than the gland in adults due to less fatty infiltration. The superficial lobe is more accessible to visualize than the deep lobe. Since the mandible and the mastoid process block the ultrasound beam, the deep lobe is partially visualized. During the scanning, the masseter muscle and mandible limit the image anteriorly; and mastoid and SCM limit the image posteriorly. Intraglandular lymph nodes are visualized as small, well-defined, and oval-shaped areas of homogenous, hypoechoic cortex with a visible hyperechoic (fatty) hilum (Fig. 10.2a) and positive Doppler’s signal (Fig. 10.2b). The unique hyperechoic hilum differentiates the node from other parotid masses. The intraglandular lymph node’s short axis in a normal state is usually less than 5–6 mm. The normal Stenson’s duct is not visible, except in slim individuals [11]. The duct becomes visualized if it is obstructed. A Doppler ultrasound quickly identifies the two blood vessels, although in some cases, the vessels are barely visible or not visible because of the high-fat content in the gland. The retromandibular vein is typically visualized as a relatively straight vessel near the center of the parotid parenchymal. The external carotid artery is visualized as a deeper vessel in Fig. 10.1e, f. The intraglandular facial nerve is not visible.

Fig. 10.2
figure 2

The characteristics of the lymph node with hyperechoic hilum (a) and positive color Doppler’s signal (b) visualized in a pediatric patient

Full size image

2.2 Submandibular Glands (SMGs)

SMGs are the second-largest SG and occupy most of the submandibular space, a potential space corresponding to the submandibular triangle. The space is formed by the superficial layer of the deep cervical fascia and is located deep to the platysma muscle and delineated by the anterior and posterior bellies of the digastric muscle and the mandible. The anterior and floor of the space is the mylohyoid muscle, and the posterior of the space is the hyoglossus muscle. The SMG may connect to the PG since it lies anteriorly to the parotid space. The gland is folded around the mylohyoid muscle’s dorsal free edge, which divides the gland to be superficial and deep lobes. Submandibular space structures include the superficial lobe of the SMG, submandibular and submental lymph nodes, facial artery, facial vein, and hypoglossal nerve’s inferior segment. Both facial artery and facial vein are the SMG vessels. The facial artery, the main blood supply, may enter the gland directly. The gland produces approximately 70% of the saliva in the unstimulated state. The Wharton duct, the primary excretory duct, exits the deep lobe’s medial surface and drains into the mouth’s anterior floor at the sublingual caruncle.

The normal SMG parenchyma appears uniformly echotexture compared with adjacent muscles (Fig. 10.3). SMGs are hypoechoic compared to the PGs and SLGs because of less fatty tissue components. The SMGs look like a triangle in longitudinal (Fig. 10.3b) and transverse (Fig. 10.3d) views. The SMGs in children appear smaller and less echogenic (darker) (Fig. 10.3e, f) than the SMGs in adults due to less fatty infiltration and may connect to the PG (Fig. 10.4a, b). The facial vein is superficial and runs along the anterosuperior part of the gland. The facial artery may cross the SMG’s parenchyma in its tortuous course and intersect the gland’s posterior pole. Wharton’s duct is not visible except in slim individuals. Lymph nodes of SMG are periglandular only.

Fig. 10.3
figure 3

Scanning of the normal submandibular gland of a pediatric patient in the longitudinal (a, b, e) and transverse (c, d, f) planes. Children have smaller and slightly darker (less echogenic) gland than adults

Full size image
Fig. 10.4
figure 4

SGUS shows the submandibular gland and parotid gland connection in an adult (a) and a 3-year-old boy (b). The submandibular glands are less echogenic than the parotid glands, especially in children

Full size image

2.3 Sublingual Glands (SLGs)

SLG is the smallest SGs in the sublingual space without a true fascial capsule. The gland lies just deep to the floor of the mouth. The SLG is superior to the mylohyoid muscle and lateral to the genioglossus and geniohyoid muscles, and medial to the mandible. The SLG size shows more significant individual variations than the PGs and SMGs. The gland has no dominant ducts and is drained by ten small ducts (the ducts of Rivinus) into the anterolateral floor of the mouth [18].

The normal SLG parenchyma appears uniformly echotexture with homogeneous hyperechogenicity compared to mylohyoid, genioglossus, and geniohyoid muscles (Fig. 10.5).

Fig. 10.5
figure 5

Scanning of the normal sublingual gland of a pediatric patient in the midline transverse plane

Full size image

3 Scanning Technique

  1. 1.

    The patient is in a supine position or upright position. If the supine position is preferred, placing a pillow or towel under the shoulders for neck extension increases thyroid, PGs, and SMGs’ accessibility and the patient/operator’s comfort.

  2. 2.

    Putting a cotton ball or a 2″ × 2″ nonsterile sponge gauze pad in each ear to protect the ultrasound gel drop into the patient’s ear canals.

  3. 3.

    Choosing an available highest-frequency probe such as a 7–14 MHz wideband linear probe. A lower frequency (5–7 MHz) probe may be used to visualize deep portions of the PGs and SMGs.

  4. 4.

    Scanning thyroid gland in a longitudinal (Fig. 10.6a, b) and transverse (Fig. 10.6c, d) planes. Thyroid’s echogenicity and homogeneity should be reviewed. Inhomogeneous features with anechoic/hypoechoic areas suggest an underlying Hashimoto’s thyroiditis (Fig. 10.6e, f).

  5. 5.

    Turning the patient’s face away from the examined side when scanning PGs and SMGs. Both PGs and SMGs should be scanned in longitudinal and transverse planes and recorded in static images and video clips.

  6. 6.

    Scanning the longitudinal plane of PGs by placing the probe parallel to the mandibular ramus and anterior to the ear (Fig. 10.1a) and scanning the transverse plane of PGs by placing the probe inferior to the ear and perpendicular to the mandible using the mandibular ramus and temporomandibular joint as landmarks (Fig. 10.1c).

  7. 7.

    Scanning the longitudinal plane of SMGs by placing the probe parallel to the mandibular body (Fig. 10.3a) and scanning the transverse plane of SMGs by placing the probe perpendicular to the body of the mandible body (Fig. 10.3c).

  8. 8.

    Scanning vessels and vascularity of PGs and SMGs with color or power Doppler.

  9. 9.

    Assessing echostructures of PGs and SMGs (Table 10.1), including echogenicity, homogeneity, hyperechoic bands, hypoechoic/anechoic areas, fatty replacement, fibrosis, calcifications, lymph nodes, posterior borders, and hypervascularity during the scanning [10].

  10. 10.

    If needed, SLGs can be scanned by placing the probe in the midline transverse or coronal plane of the submental area (Fig. 10.5a).

Fig. 10.6
figure 6

Scanning of the normal thyroid gland of a pediatric patient in the longitudinal (a, b) and transverse (c, d) planes. Inhomogeneity with anechoic/hypoechoic findings suggests Hashimoto’s thyroiditis (e, f)

Full size image
Table 10.1 Echostructures of parotid and submandibular glands [11, 21]
Full size table

4 Abnormal Salivary Gland Ultrasound Findings in Sjögren Syndrome

The essential echostructures in pSS and jSS by B-mode scanning are homogeneity, echogenicity, anechoic/hypoechoic areas. B-mode refers to a two-dimensional ultrasound image display showing bright dots of the ultrasound echoes. Echogenicity and homogeneity are the two SGUS items that determine the intraobserver and interobserver reliability [10]. The homogeneity is strongly correlated with anechoic/hypoechoic areas [11]. Both anechoic/hypoechoic spots and heterogeneous findings (cyst-like heterogeneity) are the most characteristic features for pSS and jSS (Fig. 10.7). The anechoic and hypoechoic areas possibly represent infiltration by lymphocytes, which destroy salivary parenchyma and dilated ducts [19].

Fig. 10.7
figure 7

The characteristic SGUS findings of juvenile Sjögren’s syndrome. The characteristic heterogeneity and anechoic/hypoechoic areas in parotids (a, b) and submandibular glands (c, d) of a teenager boy are shown. Similar findings are found in a 7-year-old girl with juvenile Sjögren’s syndrome (e, f). An enlarged intraglandular lymph node (a) and hyperechoic bands (a, b) are noted

Full size image

Other SGUS findings (Table 10.1) are hyperechoic foci or intraglandular calcifications, hyperechoic bands, fibrosis, enlarged lymph nodes (Fig. 10.7a), fatty replacement, and changes of the posterior border and the gland size [11, 20, 21]. The hyperechoic calcified foci are typically tiny and do not demonstrate a posterior acoustic shadow. Hypervascularity assessed by color or power Doppler indicates acute flare or SG parenchyma inflammation (Fig. 10.8). The abnormal echostructures do not significantly change within the 2 years of scanning in pSS [22]. The SGUS findings are not pathognomonic for pSS and jSS. Additional clinical information is required to interpret the SGUS abnormalities.

Fig. 10.8
figure 8

Hypervascularity of active inflammation of the parotid gland (a) and submandibular gland (b) assessed by color Doppler

Full size image

SGUS findings of the right major SG are correlated with the left major SG in pSS. If both PG and SMG scanning cannot be completed, at least one PG and one SMG should be scanned in pSS [23]. It is unknown whether the correlation between the right and left major SG is the same as that of the SGUS findings in jSS. It is possible that the findings of PG are not correlated with those of SMGs since those two types of major SGs glands do not share the same parenchymal echostructure, and the disease severity affecting the glands can vary.

5 Salivary Gland Ultrasound Scoring Systems

SGUS has clinically proved its value for the evaluation of pSS and jSS. Various semiquantitative SGUS scoring systems for the diagnosis of pSS were proposed based upon the B-mode scanning. The two most reliable and consistent scanning results, echogenicity and homogeneity, are incorporated in most scoring systems. Hyperechoic bands, calcifications, glandular borders, and lymph nodes are unreliable and difficult to assess; however, these findings are applied to certain scoring systems [11, 21].

The first semiquantitative scoring system of SGUS was developed in 1992 [24]. Over 20 scoring systems were subsequently developed in the past two decades. The scoring systems are classified into three systems based upon their total scores: 0–4, 0–16, and 0–48. The Hocevar 0–48 scoring system is the most comprehensive and was applied to several pSS studies [25]. The 0–4 scoring systems show higher specificity and less heterogeneity than others. The 0–4 scoring system is recommended as a universal diagnostic standard [9].

5.1 Salivary Gland Ultrasound Scoring System and the Diagnosis of pSS/jSS

A meta-analysis of SGUS scoring systems in diagnosing pSS showed a pooled sensitivity of 69% and specificity of 92% [8]. A negative SGUS result does not exclude a diagnosis of pSS. Table 10.2 shows a few scoring systems with details and the cutoff. Implementing the scoring systems into the 2016 ACR/EULAR classification criteria for the diagnosis of pSS increases the criteria’s sensitivity without changing its specificity (Table 10.3). Recently, a novel, reliable, semiquantitative scoring system of SGUS was developed as a consensual scoring system [11

Table 10.2 Semiquantitative salivary gland ultrasound scoring systems
Full size table

]. The usefulness of the novel scoring system in diagnosing pSS and jSS requires further studies.

Table 10.3 Incorporation of salivary gland ultrasonography into the 2016 ACR/EULAR classification criteria
Full size table

5.2 Salivary Gland Ultrasound Scoring System and Salivary Gland Biopsy

Abnormal SGUS is correlated with minor salivary gland biopsy (MSGB) and PG biopsies in pSS [26,27,28,29,30]. Although a positive SGUS is highly predictive of positive biopsies, they are not interchangeable. The SGUS result cannot replace MSGB in the criteria. Approximately, 16 and 22% of positive SGUS have negative MSGB and PG biopsies, and 26 and 14% of negative SGUS have positive MSGB and PG biopsies accordingly [31]. MSGB remains necessary to perform if SGUS is negative. If SGUS replaces MSGB in the 2016 ACR/ EULAR classification criteria, the criteria’s sensitivity is substantially reduced.

5.3 Salivary Gland Ultrasound Scoring System and Autoantibodies

Abnormal SGUS may predict SS antibodies’ positivity in pSS because of the good agreement of the SGUS and SS antibody results [7, 27, 31,32,33]. The majority of cases with positive SGUS and anti-SSA fulfill the 2016 ACR/EULAR classification criteria [7, 31]. However, more studies are needed to support whether the combination of the positive SGUS and anti-SSA can replace the MSGB result. More studies of the positive SGUS and anti-SSA are needed to support the findings.

5.4 Salivary Gland Ultrasound Scoring System and Sialometry

Abnormal SGUS show an inversed correlation with unstimulated whole salivary flow studies [7, 26,27,28, 30]. However, the accuracy of abnormal SGUS to predict salivary gland dysfunction is known to be poor [31].

6 SGUS Mimicking Sjögren’s Syndrome

6.1 Juvenile Recurrent Parotitis (JRP)

Juvenile recurrent parotitis (JRP) is an unknown cause of recurrent nonobstructive, nonsuppurative parotitis in children with unilateral or bilateral PG swellings. The typical presentation is pain, swelling, redness of skin overlying the gland, and fever (Fig. 10.9a). Chronic recurrent parotitis or chronic sialadenitis in adults with the same clinical and sonographic features is less common [34]. SGUS findings in JRP are multiple anechoic/hypoechoic areas, heterogeneous parenchyma echostructures, and hypervascularization with increased Doppler’s signal [35]. The hypoechoic areas in JRP correspond to either sialectasis or lymphocytic infiltration [36]. Both JRP and jSS share the same clinical and PG sonographic findings (Figs. 10.9 and 10.10), especially in young children with parotitis and negative SSA/SSB antibodies. Since JRP does not involve minor SGs and SMGs, the positive MSGB and abnormal SMG ultrasound may distinguish these two conditions. Magnetic resonance imaging (MRI) findings of PGs without fat degeneration and architecture destruction possibly suggest JRP [37].

Fig. 10.9
figure 9

A 6-year-old girl presented with recurrent pain, swelling, and redness of the right parotid gland (a) and SGUS revealed heterogeneous parenchyma with multiple anechoic/hypoechoic areas (b). Both juvenile Sjögren’s syndrome and juvenile recurrent parotitis are common differential diagnoses in young children

Full size image
Fig. 10.10
figure 10

SGUS findings of chronic bacterial infection in a teenager with recurrent pain and swelling of the parotid gland

Full size image

6.2 Mumps

Mumps, a paramyxovirus infection, is characterized by painful and inflamed PGs and constitutional symptoms. Bilateral PG involvement is shown in 70% of cases, PG and SMG involvement is presented in 11% of cases, and isolated SMG involvement is noted in 1–2% of cases. Although 30% of infected cases are asymptomatic, severe complications including orchitis, oophoritis, mastitis, pancreatitis, aseptic meningitis, encephalitis, and deafness were reported. SGUS findings in mump are enlarged SG, homogenous echogenicity, heterogeneously hypoechoic intraglandular lymph nodes, and hypervascularity [35, 38].

6.3 Chronic Bacterial Infection

An incomplete resolution of an acute infection caused by Staphylococcus aureus, Streptococcus spp., anaerobic bacteria, and aerobic and facultative Gram-negative bacilli leads to chronic bacterial infection [39]. SGUS shows multiple oval hypoechoic areas, varying degrees of sialectasis, and hyperechoic septations with hypervascularization (Fig. 10.10) [40].

6.4 IgG4-Related Disease

IgG4-related disease (IgG4-RD), a fibro-inflammatory disorder, commonly involves SGs [41]. Classic findings of IgG4-RD are bilateral symmetric diffuse swelling of SGs and lacrimal glands. Ultrasonography is considered the most useful imaging modality for IgG4-RD. The characteristic of ultrasound findings in IgG4-RD is a nodal pattern predominantly affecting SMGs (92%) with less-affecting PGs (33%) [42]. The nodal pattern is described as hypoechoic areas with high vascularization and bulging from normal surface commonly detected in SMGs. Both SMGs and PGs are typically homogenous with normal parenchymal echogenicity. The changes are not observed in pSS. The nodal pattern detected by SGUS may not be seen on CT and MRI [43].

6.5 Sarcoidosis

Sarcoidosis, a chronic multisystem inflammatory granulomatous disease with unknown etiology, affecting the SGs, involves mostly PGs. Both sarcoidosis and pSS may have similar signs and symptoms with persistent asymptomatic or painful PG swelling. SGUS findings of sarcoidosis are multiple hypoechoic areas and numerous echogenic septa with the hypervascular area and heterogeneous parenchyma with or without enlarged intraglandular lymph nodes [44]. The findings in sarcoidosis cannot distinguish from the findings in pSS [45].

6.6 Systemic Immunoglobulin Light-Chain Amyloidosis (AL Amyloidosis)

AL amyloidosis is the most common form of systemic amyloidosis characterized by the proliferation of clonal plasma cells and immunoglobulin light-chain production, leading to SG enlargement and other organ damage [46]. SGUS findings in AL amyloidosis are hypoechoic lesions and hyperechoic septa, similar to the findings in pSS [45].

6.7 HIV Infections

HIV-associated SG disease commonly involves bilateral PGs. Typically, the parotitis is asymmetrical, painless, slow-growing with xerostomia, and salivary gland hypofunction. The PG ultrasound findings in HIV-infected patients can be normal and abnormal. There are four distinct ultrasound patterns including (1) lymphocytic aggregations (diffuse small hypoechoic or anechoic areas with moderate to ill-defined margins interspersed within normal isoechoic areas, not associated with posterior acoustic enhancement), (2) lymphoepithelial cysts, a pathognomonic of HIV infection (well-circumscribed round hypoechoic area with size 5 mm or larger and posterior acoustic enhancement), (3) fatty infiltration (whole hypoechoic gland), and (4) lymphadenopathy (oval-shaped hypoechoic areas or nodules with echogenic hilum and hilar blood flow). Both lymphocytic aggregations and lymphoepithelial cysts may be similar to the findings in SS [47].

6.8 Lymphoma

Lymphomas are the second most common malignancies in the head and neck. SG lymphomas can be primary or secondary, and extranodal (arising outside a lymphoid organ) or nodal (appearing inside an intraglandular lymph node or periparotid lymph node) [48]. Primary SG lymphomas commonly involve unilateral or bilateral PGs, and most are classified as extranodal marginal-zone B-cell non-Hodgkin’s lymphoma originating from mucosa-associated lymphoid tissue (MALT) [49]. SS is a well-known risk factor for primary SG lymphoma. There are two defined sonographic patterns of SG lymphomas, including linear echogenic strands pattern and segmental pattern. The linear echogenic strands pattern is a marked hypoechoic area with interspersed linear echogenic strands, and the segmental pattern is multiple, relatively large hypoechoic segments. The SGUS findings represent the histopathologic findings of expanding lymphoma cells demarcated by narrow or wide fibrous bands [50, 51].

6.9 Postradiotherapy

Radiotherapy alters and damages PG parenchyma and SG acinar cells leading to xerostomia. SGUS of PG postradiotherapy appears isoechoic or hypoechoic echogenicity, small gland size, heterogeneous echotexture with multiple hypoechoic areas, and hyperechoic fibrotic streaks [52].

7 Other Imaging for Primary Sjögren’s Syndrome and Juvenile Sjögren’s Syndrome

7.1 Plain Radiograph

A plain radiograph is an inexpensive and simple study for evaluating salivary stones, especially SMG stones since the SMG stones are relatively large and radiopaque [53]. It has no diagnostic role for pSS and jSS.

7.2 Sialography

Sialography is an old diagnostic procedure based on the injection of a dye into the salivary duct orifice (Stenson and Wharton duct) to visualize the entire ductal architecture and identify salivary calculi, ductal anomaly, and stricture. Abnormal sialography with diffuse sialectasis, including punctate, cavity, or destructive pattern, is a criterion in the 2002 American-European Consensus Group (AECG) classification. The abnormal findings result from cystic ductal dilatations or extravasation of the contrast material into the parenchyma. The extravasation of the contrast material is possibly related to the dysfunction of tight junctions between striated ductal cells caused by inflammatory cytokines. Its sensitivity and specificity are 80 and 89% based upon the 2002 AECG criteria [54].

Sialography is excluded from the 2016 ACR/EULAR criteria because of its invasiveness, potential complications including ductal rupture or activation of dormant infection, irradiating nature, and contraindications such as acute infection, acute inflammation, and contrast allergy. SGUS has replaced sialography since it provides similar information while being considerably less invasive.

7.3 Salivary Gland Scintigraphy

Salivary gland scintigraphy, a nuclear imaging technique with radioactive tracer infusion (Technetium-99m pertechnetate), demonstrates the glandular function by evaluating the distribution and speed of eliminating the radiotracer after a secretive stimulation. It reveals both physiologic and pathologic secretory functions. The potential benefit of scintigraphy is to monitor SG functioning over time.

Abnormal scintigraphy results are delayed uptake, reduced concentration, and delayed excretion of tracer described in the 2002 AECG criteria. Its sensitivity and specificity are 89 and 50% [55]. Although it is noninvasive, safe, well-tolerated, reproducible, and easy to perform, the test results cannot distinguish the secretive dysfunction of pSS from other diseases [55]. In addition, there are no guidelines or consensus on how the test should be performed and interpreted [56].

7.4 Computed Tomography (CT)

Computed tomography (CT) and MRI identify the homogeneity and inhomogeneity of adipose tissue distribution and accumulation related to destruction and shrinkage from chronic inflammation. CT is more readily available and lower-cost with a shorter scan time than MRI. Non-contrast-enhanced CT is useful in identifying small calculi in the gland or duct. Contrast-enhanced CT is useful where MRI is contraindicated, or the access to MRI is limited.

Normal fatty tissue inside the PGs is homogeneously distributed and contributes to the low CT attenuation regardless of the subject’s age. Abnormal CT findings of the PGs in SS include heterogeneity of attenuation caused by fatty degeneration, abnormal fat tissue deposition, diffuse punctate calcification, atrophy or swelling of the gland, nodularity, and cystic lesions. The diffuse punctate calcification of bilateral PGs is highly specific for SS and found in 30% of cases [57]. The drawback of CT scan is radiation exposure, artifacts from metal dental fillings, and its contraindications, including impaired kidney function and contrast hypersensitivity.

7.5 Magnetic Resonance Imaging (MRI) and Magnetic Resonance Sialography (MR Sialography)

MRI is noninvasive, nonionizing imaging, providing pathologic changes of major SG in SS. T1-weighted image and fat-suppressed T2-weighted image are recommended. Intravenous injection of a contrast medium increases the detection of abnormalities, but it is not mandatory for SS diagnosis. Normal PGs are intermediate to high single intensity on T1- and T2-weighted images due to high intraglandular fat content. The submandibular and sublingual glands demonstrate a more intermediate signal on T1-weighted image [58]. MRI can identify early SS with edema and gland enlargement, lobular destruction with diffuse micro and/or macrocystis changes associated with deposition of fibrous tissue and fat [55]. Both high-signal intensity on T1- and T2-weighted images and low-signal intensity in fat-suppressed T2-weighted images indicate abnormal adipose tissue deposition. The characteristic MRI of PGs and SMGs is heterogeneous signal intensity (mixed low- and high-signal intensity) distribution on T1- and T2-weighted images with a salt-and-pepper appearance or a honeycomb appearance (Fig. 10.11).

Fig. 10.11
figure 11

A characteristic MRI finding of the parotid gland of a teenager with juvenile Sjögren’s syndrome shows salt-and-pepper appearance (a, white arrow) compared to SGUS (b)

Full size image

MR sialography, a new modification of MRI, reveals early diseases of PG without the injection of contrast material. MR sialography is obtained with a heavily T2-weighted 3D-Fast spin-echo appearance. MR sialography’s abnormal findings are multiple high-signal-intensity spots with apple tree appearance and enlargement or reduction of the gland’s size. A combination of MRI and MR sialography provides high sensitivity and specificity for observing parenchymal abnormalities and differentiation of SS [59, 60].

The significant disadvantages of MRI and MR sialography are higher cost, length of the examination, patient intolerance, sedation requirement in young children, and its contraindications, including retained metallic foreign bodies such as dental braces, cardiac pacemakers, or insulin pumps.

8 Conclusion

SGUS has been increasingly used as a first-line imaging tool to evaluate and diagnose pSS and jSS with reasonable accuracy. It offers a noninvasive assessment of typical morphological changes in the major SGs. There is growing evidence to support the abnormal sonographic features as a criterion for the next disease classification. The potential roles of SGUS in reducing the need for MSGB and monitoring treatment response require further studies.