The Egyptian Rheumatologist (2011) 33, 147–154

Egyptian Society for Joint Diseases and Arthritis

The Egyptian Rheumatologist

www.rheumatology.eg.netwww.sciencedirect.com

ORIGINAL ARTICLE

The value of magnetic resonance sialography and magnetic

resonance imaging versus conventional sialography

of the parotid gland in the diagnosis and staging

of Sjogren’s syndrome

Nashwa S. Ahmed a, Soad M. Mansour a, Mohamed M. El-Wakd b,*,

Hatem M. Al-Azizic, Noha S. Abu-Taleb

a

a Oral Radiology Department, Faculty of Oral and Dental Medicine, Cairo University, Egyptb Rheumatology and Rehabilitation Department, Faculty of Medicine, Cairo University, Egyptc Radiology Department, Faculty of Medicine, Cairo University, Egypt

Received 4 March 2011; Accepted 23 May 2011Available online 17 June 2011

*

E-

11

Pr

Pe

an

do

KEYWORDS

Sjogren’s syndrome;

Parotid gland;

Conventional sialography;

Magnetic resonance sialog-

raphy;

Magnetic resonance imaging

Corresponding author.

mail address: elwakd@yahoo

10-1164 � 2011 Egyptian So

oduction and hosting by Els

er review under responsibility

d Arthritis.

i:10.1016/j.ejr.2011.05.004

Production and h

.com (M

ciety for

evier B.V

of Egyp

osting by E

Abstract Introduction: The conventional sialography (CS) and minor salivary gland biopsy have

been considered the cornerstones of the diagnosis of Sjogren’s syndrome (SS). However, they are

invasive and with risk of complications to the patient. Magnetic resonance imaging (MRI) has been

widely replacing conventional invasive examinations. They provide insufficient data for diagnosing

and staging SS, because while they address the detection and diagnosis of parenchymal salivary

gland disease they do not assess duct abnormalities. The introduction of magnetic resonance sialog-

raphy (MRS) protocol was the first step towards non-ionizing non-invasive technique enables a

more accurate assessment of the duct system. Moreover, it is postulated that the combined use

of MRI and MRS is useful for the assessment of damaged salivary glands in SS.

.M. El-Wakd).

Joint Diseases and Arthritis.

. All rights reserved.

tian Society for Joint Diseases

lsevier

148 N.S. Ahmed et al.

Aim of work: This preliminary study aimed to assess the role and efficacy of MRS and MRI in

imaging the parotid gland in SS patients’ and to compare their results with that of CS in the

diagnosis and staging of SS.

Patients and methods: The parotid gland was examined by MRS, MRI and CS in 15 SS patients.

Scoring system for overall branching pattern, ductal system staging, sialectasis for both CS and

MRS was used. MRI scoring system for glandular parenchymal pattern, size and contour and

lymph node was used.

Results: This revealed non-significant difference between CS and MRS findings in the diagnosis

and staging of SS. Furthermore, no statistically significant correlation was found between the clin-

ical and SS stage in both CS and MRS. On the other hand, a statistically significant positive cor-

relation was found between the clinical findings and the MRI parenchymal stages. Moreover,

MRS showed higher sensitivity and diagnostic accuracy values than CS. Combining the MRS

and MRI abnormalities, 100% sensitivity diagnostic accuracy values were achieved in the diagnosis

and staging of SS. The inter-observer agreement was higher in MRS than in CS and was perfect in

MRI.

Conclusion: In this preliminary study, MRS showed a higher sensitivity and accuracy in diagnos-

ing and staging SS and may safely and securely replace CS. MRI and MRS give information on

different aspects of glandular and duct pathology; therefore, both should be performed when exam-

ining the parotid glands in SS patients.

� 2011 Egyptian Society for Joint Diseases and Arthritis. Production and hosting by Elsevier B.V.

All rights reserved.

1. Introduction

Sjogren’s syndrome (SS) is a chronic systemic autoimmunedisease characterized by lymphocytic infiltration of exocrine

glands particularly the salivary and lacrimal glands interfer-ing with the normal gland function resulting in dry mouthand eyes [1,2]. The diagnosis is based on the combined eval-uation of multiple clinical, serological, functional, histopa-

thological and morphological parameters [2–4]. Minorsalivary gland biopsy and conventional sialography (CS) havebeen considered the cornerstones of the diagnosis of SS [5,6].

However, they are both invasive examinations that may causeinconvenience and a risk of complications, such as dye reten-tion, inability to duct cannulation and intra- or post-proce-

dure pain [7,8]. Ultrasonography and magnetic resonanceimaging (MRI) have been widely replacing conventional inva-sive examinations. They provide insufficient data for accu-rately diagnosing and staging SS, because while they

address the detection and diagnosis of parenchymal diseasesof salivary glands they do not assess salivary gland ductabnormalities [8–12]. The introduction of magnetic resonance

sialography (MRS) protocol was the first step towards non-ionizing non-invasive technique enables a more accurateassessment of the duct system [13]. Moreover, it is postulated

that the combined use of high resolution MRI, by T1weighted or fat-suppressed T2-weighted imaging, and MRSis useful for the assessment of damaged salivary glands in

SS. However, careful validation of MRI findings is requiredbefore the importance of these findings is established with re-spect to disease activity. If properly validated, high resolutionMRI of the salivary glands may become a powerful noninva-

sive test for examining and staging damages to the glands inSS [14], and for differentiation from other diseases that mimicSS [8,15]. This preliminary study aimed to assess the role and

efficacy of MRS and MRI in imaging the parotid gland in SSpatients’ and to compare their results with that of CS in thediagnosis and staging of SS.

2. Patients and methods

2.1. Patients

Fifteen female SS patients were selected from the Rheumatologyand Rehabilitation Department, Faculty of Medicine, CairoUniversity according to the revised version of the European cri-

teria proposed by the American-European Consensus Group[3]. They were subjected to thorough history taking and clinicalexamination both systemic and local to the parotid gland area.

2.1.1. MRI examination

For each patient, either right or left parotid gland wasexamined by using a 1.5-T superconductive magnet (Philips;

Gyroscan Intera; Netherlands). A surface coil was used. AxialT1-weighted images (T1WI) and axial STIR (short time inver-sion-recovery) images were performed. T1WI: was performed

with the following parameters: Repetition time (TR): 525, echotime (TE): 13, number of signal averages/excitations (NSA/NEX): 8, field of view (FOV): 7 cm, image matrix: 256 · 256,

slice thickness: 2 mm, inter-slice gap: no gap, imaging time:3 min and 46 s, STIR images: were performed with the follow-ing parameters: TR: 2650, TE: 90, NSA/NEX: 8, FOV: 7 cm,image matrix: 256 · 256, slice thickness: 2 mm, inter-slice gap:

no gap and imaging time: 5 min.

2.1.2. MRS

Sagittal MRS images were performed using single section sin-gle-shot turbo spin-echo (SSTSE) sequence with the followingparameters: TR: 8000, TE: 900, NSA/NEX: 8, FOV: 7 cm,image matrix: 256 · 256, slice thickness: 3 cm and imaging

time: 48 s. A rectangular image area for MRS was deter-mined on the axial STIR images and its long axis was placedparallel to the course of the main duct. The slice thickness

was thick so that the whole gland area was included in theimage area.

The value of magnetic resonance sialography and magnetic resonance imaging 149

2.1.3. CS

All patients underwent CS after MRI and MRS examinations

by using the angiography system (Diagnost 94; Philips;Holland).

2.1.4. Scoring system for CS findings

The normal parotid sialogram shows continuous branching ofthe ductal system tapering off towards the periphery of theglands, smooth contours of the main duct and ductal branches

and no sialectasias. A rarified ductal system, irregular or de-stroyed ductal contours, occlusions and sialectasiaswere consid-ered pathologic findings [16]. Overall branching pattern:

Normal: 0. Sparse: 1. The width and contour of Duct system:(main duct, primary and secondary ductal branches): Normal:0, narrowwith or without irregularity: 1, dilated with or without

irregularity: 2, and destroyed: 3. The sialectasis: based on thewidely accepted criteria proposed by Rubin and Holt [17].According to the size of contrast material collections inside thegland: Stage-0:No sialectasis. Stage-I: Punctate;<1 mm indiam-

eter. Stage-II:Globular; 1–2 mm indiameter. Stage-III: Cavitary;>2 mm in diameter. Stage-IV: Destructive; severe irregular dila-tation of the main duct and destruction of the parenchyma with

pooling of contrast medium giving a bizarre pattern. Diagnosisof SS was made when the findings of CS showed stage 1 diseaseor greater. According to the number of sialectasia: Absent: 0.

Few: 1. Many: 2. Each of accessory duct/gland, filling defectand stricture were given 0 if absent and 1 if present.

2.1.5. Scoring system for MRS findings

The same scoring system of CS findings was applied to theMRSfindings; where the MRS staging was determined according tothe size of the high signal intensity areas inside the gland.

2.1.6. Scoring system for MRI findings

2.1.6.1. Glandular parenchymal pattern. It was relied on T1WI

and STIR images. This was a ‘‘Modification’’ to the staging sys-tem of Takagi et al, 2005 [2]. Stage-0: Normal, the parotid glandis of homogenous intermediate signal intensity denoting intact

gland parenchyma. Stage-I: Mild distribution of heterogeneousparenchymal pattern (<25%of the total gland area)within nor-mal parenchyma. Stage-II: Moderate distribution of heteroge-

neous parenchymal pattern (25–50% of the total gland area)within normal parenchyma. Stage-III: Advanced distributionof heterogeneous parenchymal pattern (50–75 % of the total

gland area) within normal parenchyma. Stage-IV: Severe, thereis a diffused distribution of heterogeneous parenchymal pattern(>75% of the gland area). Gland size: Normal: 0. Enlarged: 1.Gland contour: Smooth: 0. Irregular: 1. Accessory gland: Ab-

sent: 0. Present: 1. Lymph nodes: Absent: 0. Present: 1.

2.1.7. Image analysis

MRI, MRS and CS were evaluated independently by threeradiologists. The findings of CS were used as the standard of ref-erence. Four-point scale for the evaluation of the image qualityregarding clarity, visibility and resolution to determine the inter-

observer agreement in each imaging modality: 0 = Poor images.1 = Fair images. 2 = Good images. 3 = Very good images.

The current study design was discussed and approved by

the Selected-Review-Board, Faculty of Oral and Dental Med-icine, Cairo University, on May 2007. A written informed con-sent was obtained from all the patients sharing in the study.

Statistics: Qualitative data were presented as frequenciesand percentages. Chi square test was used for comparisonsbetween the variables. Yates’ correction was used to prevent

overestimation of statistical significance for small data. Thisformula was chiefly used when at least one cell of the tablehad count smaller than five. Quantitative data were pre-

sented as means and SD values. Spearman’s correlation coef-ficient was used to determine significant correlations betweendifferent variables. The significance level was set at P 6 0.05.

Statistical analysis was performed with SPSS (StatisticalPackage for Scientific Studies) 16.0� for Windows. Sensitiv-ity, Specificity and Diagnostic accuracy were calculated asfollows:

Sensitivity ð%Þ ¼ True positive

True positiveþ false negative� 100

Diagnostic accuracy ð%Þ ¼ True positiveþ True negative

Total number

� 100

Kappa statistic was used to measure the agreement betweenthe evaluations of the different raters (Inter-observer agree-

ment). A value of 1 indicates perfect agreement; it is zero whenthere is no agreement. Kappa values can be interpreted as fol-lows: poor (<0.2), fair (0.21–0.4), moderate (0.41–0.6), good

(0.61–0.8) and very good (0.81–1.00).

3. Results

The patient’s age varied from 30 to 57 years with mean of42.5 ± 10 years with disease duration varied from 2 to15 years with a mean of 6.1 ± 3.8 years. All of them have

keratoconjunctivitis sicca and 14 have Extra-glandular mani-festations (93.3%). Recurrent parotid gland swelling wasnoted in 9 (60%) patients.

3.1. Comparison between CS and MRS findings

The findings between CS and MRS is demonstrated in Table 1and Figs. 1–4.

3.2. MRI findings

Glandular parenchymal pattern: stage-0 was found in one

(6.7%) patient, stage-I in 7 (46.7%) patients; stage-II, IIIand IV in 3 (20%) patients for each. Gland size: 7 patients(46.7%) had enlarged parotid gland. Gland contour: 7 patients

(46.7%) had irregular parotid gland. Other findings: accessorygland was detected in 12 (80%) patients and lymph nodeenlargement in 11 (73.3%) patients. Different patterns wereshown in fig. 5.

3.3. Laboratory findings

Positive anti Ro/La antibodies were found in 11 (73.3%)

patients.

Table 1 Comparison between CS and MRS findings.

Findings Group P-value

CS (n= 14) MRS (n= 15)

Frequency % Frequency %

Overall branching pattern

Normal 5 35.7 4 26.7 0.6

Sparse 9 64.3 11 73.3

The duct system

Main duct Normal 5 35.7 6 40 0.984

Narrow 2 14.3 2 13.3

Dilated 7 50 7 46.7

1ry branches Normal 6 42.9 7 46.7 0.979

Narrow 1 7.1 1 6.7

Dilated 7 50 7 46.7

2ry branches Normal 7 50 3 20 0.192

Narrow 2 14.3 5 33.3

Dilated 2 14.3 2 13.3

Destroyed 3 21.4 5 33.3

Sialectasis

Size Stage-0 No sialectasis 3 21.4 2 13.3 0.7

Stage-I Punctate 7 50 8 53.3

Stage-II Globular 3 21.4 3 20

Stage-III Cavitary 1 7.1 2 13.3

Stage-IV Destructive 0 0

Number Absent 3 21.4 2 13.3 0.66

Few 6 42.9 10 66.7

Many 5 35.7 3 20

Other findings

Accessory gland/duct 7 50 12 80 0.09

Filling defect 2 14.3 0 0 0.13

Stricture 6 42 6 40 0.59

Figure 1 Overall branching pattern of the parotid gland by CS;

normal (left) and sparse (right).

150 N.S. Ahmed et al.

3.4. Correlation analysis using Spearman’s correlationcoefficient test

Ductal stage and SS stage (sialecatic stage) in both CS andMRS showed no statistically significant correlation (P >0.05). MRI parenchymal stage and MRS SS stage (sialecatic

stage) showed no statistically significant correlation (P >0.05). MRI parenchymal stage and MRS ductal stage showeda statistically significant positive correlation between MRI

parenchymal stage and 1ry & 2ry branches (P = 0.04 and0.03 respectively). SS stage (sialecatic stages) and age, disease

duration, clinical and laboratory findings in both CS andMRS showed no statistically significant correlation. Mainductal stage in both CS and MRS and age, disease duration,

clinical, laboratory findings in both CS and MRS showed nostatistically significant correlation. There were a statisticallysignificant positive correlation between age (P< 0.001), dis-ease duration (P = 0.004), recurrent parotid swelling (P =

0.027) and MRI parenchymal pattern. There was a statisticallysignificant positive correlation between anti Ro/La antibodiesand parenchymal pattern (P = 0.02). There was no statisti-

cally significant correlation between parenchymal patternfindings and extra-glandular manifestations.

3.5. Diagnostic performance of the imaging modalities

MRS showed higher sensitivity and diagnostic accuracy values(86.7% for each) than CS (78.6% for each) when the presence

of sialectasis was taken as the reference for the diagnosis of SS.Moreover, MRS showed slightly higher sensitivity and diag-nostic accuracy (93.3% for each) than CS (92.9% for each)when any abnormality of the ductal system and sialectasis were

considered positive for SS. As regard the MRI findings, theparenchymal stage has high sensitivity and diagnostic accuracyvalues (93.3% for each), however diagnosis based on gland size

or contour yield a very low sensitivity and accuracy values(46.7% for each). Combining MRS and MRI revealed 100%sensitivity and 100% diagnostic accuracy in the diagnosis of

SS.

Figure 2 Duct system. CS showed (A) stage-0, (B) stage-I and

(C) stage-II. MRS showed (D) stage-0, (E) stage-I and (F) stage-II.

Figure 3 Sialectasis. CS showed: (A) SS stage-I, (B) SS stage-II, (C)

stage-III.

Figure 4 CS (left) and MRS (right). Accessory parotid gland

(black star), filling defect (black arrow head) and main duct

stricture (white star).

The value of magnetic resonance sialography and magnetic resonance imaging 151

3.6. Inter-observer agreement using Kappa statistic

As regard CS, there was a statistically significant agreementbetween all observers (P< 0.001) as regards the main duct;the agreement ranged from moderate (0.573) to good (0.632).

As regards the secondary branches, significant agreement(P< 0.001) was only found between Observer 2 and observer3; the agreement was good (0.712). There was non-statistically

significant agreement as regards the overall branching pattern,1ry branches and sialectasis.

As regard MRS, there was a statistically significant perfect

(1.00) agreement between all observers (P < 0.001) as regardsthe main duct. Moreover, there was a statistically significantagreement between all observers (P < 0.005) as regards the

sialectasis; the agreement ranged from moderate (0.559) tovery good (0.865). As regards the overall branching pattern,significant good (0.745) agreement was only found between ob-server 2 and 3. As regards the primary branches, significant

moderate (0.471) agreement was found between Observer 1and 3. The same agreement happened between Observer 2and 3. While; for secondary branches; significant moderate

(0.559) agreement was only found between Observer 2 and 3.

SS stage-III. MRS showed: (D) SS stage-I, (E) SS stage-II, (F) SS

Figure 5 Parotid gland parenchymal pattern (above); T1WI showed (left to right) normal homogenous gland, mild, moderate, advanced,

and severe heterogenous gland pattern. Parotid gland contour (below); STIR image showed regular (left) and irregular gland contour

(middle). Lymph node enlargement (right).

152 N.S. Ahmed et al.

As regard MRI, there was a statistically significant agree-

ment between all observers (P < 0.001) as regards parenchy-mal pattern and gland contour where the agreement wasperfect (1.00).

4. Discussion

This preliminary study aimed to assess the role and efficacy of

MRS and MRI in imaging the parotid gland in SS patients’and to compare their results with that of CS in the diagnosisand staging of SS.

According to our results, both modalities (MRI/MRS andCS) showed similar stages of the main and primary ductalbranches. The main duct was normal in 35.7% in CS versus40% in MRS (where CS was technically failed in one patient

whose main duct was normal in MRS); narrow in 14.3% inCS versus 13.3% in MRS and dilated in 46.7% in CS versus50% in MRS. No destroyed main ducts. Regarding the pri-

mary ductal branches, both modalities showed similar ductalstages. However, Stiller et al. [16] found that 56.5% of sialo-grams revealed normal main ducts, 34.8% revealed narrow

main duct and 8.7% revealed dilated ducts. No destroyed mainducts were observed. Additionally, MRS findings of Niemelaet al. [18] revealed that 3.8% of patients had normal ducts;

65.4% had narrow ducts and 30.8% had dilated ducts. Inthe present study, the secondary ductal branches were narrowor destroyed in 35.7% in CS versus 66.6% in MRS; yet, thesedifferences were statistically non-significant. Taking CS as the

gold standard reference, the higher percentage of narrowing inMRS may be due to the non-use of sialogogue in our studyresulting in improper visualization of the minor ductal

branches. Ductal structures were delineated more clearly afterstimulation with a sialogogue [19,20]. Furthermore, CS offershigher spatial resolution than MRS in the assessment of sec-

ondary and tertiary ductal branches; however this was not very

important as the diagnosis of the disease is often not affected[21,22].

Diagnosis of SS depends upon the sialectasis and the stage

is determined according to its size, which is important in boththe initial diagnosis and the follow-up as it has an effect on thetreatment plan and on the prognosis [23]. In the current study,

the SS stage-0 was found in 21.4% in CS versus 13.3% inMRS, however, stage-1 was observed in 50% in CS versus53% in MRS; this was due to one patient was judged as stage0 in CS and stage 1 in MRS. This finding may be attributed to

the severe pain experienced by the patient during injectionwhich was stopped immediately afterwards resulting in incom-plete filling of the ductal system, thus sialectasis were not

apparent in CS. On the contrary, Takagi et al. [8] in their studyfound that some patients with grade 0 in MRS showed grade 1changes in CS. Stage-2 was observed in 21.4% in CS and 20%

in MRS; stage-3 was observed in 7.1% in CS versus 13.3% inMRS. These results came in agreement with the results of laterstudy who found that MRS findings were well correlated withCS in stages 2 and 3 of the disease8. In our study, no stage-4

was found among our patients. Despite the previous figures,our results showed no statistically significant differencesbetween the CS and MRS regarding the disease staging. More-

over, in 86.6% of patients the stage of SS determined withMRS was identical to that determined with CS. Almost nearto our results; Ohbayashi et al. [24] found that CS stages

and MRS stages were well correlated in 89% of their patients.Also, Yu CQ et al. [15], in their study found that 83.3% of pa-tients have the same manifestations by degree classification.

However, Tonami et al. [23] in their study had found similarstages in 50% of patients.

With regards to the number of sialectasis, our study re-vealed many sialectasis in 35.7% patients in CS versus 20%

The value of magnetic resonance sialography and magnetic resonance imaging 153

in MRS. This finding suggests that CS and MRS depict differ-ent aspects of the disease where in CS sialectasis representdilated intraglandular ducts besides areas of pooled contrast

material due to fragmentation of the duct walls. However,MRS is a hydrography technique that represents the water richstructures only. The non-significant differences established be-

tween CS and MRS findings and staging of SS, more to thepoint the complications of CS, all these suggested that MRScan be safely and securely replace the invasive CS in the diag-

nosis and staging of SS.According to MRI results, T1WI revealed high signal inten-

sity areas throughout the gland which were changed to low sig-nal intensity on STIR images. Four stages were suggested.

These changes yielded a heterogeneous parenchymal patternin 93.3% patients; where stage-I was found in 46.7% patients,stage-II in 20% patients, stage-III in 6.1% patient and stage-

IV in 20% patients. In accordance with our results, Izumiet al. [25] noticed that the parotid glands in SS showed heter-ogeneity of signal intensity on MRI and believed that this may

indicate the presence of hemorrhage, proteinaceous content, orfat tissue. In 1997 [26], they performed another study andfound that the high-intensity signals in T1WI which were sup-

pressed in STIR images proving that these areas definitelyareas of fat deposition. Additional study found that glands af-fected with SS were heterogeneous and consisted of low andhigh signal intensity areas (representing fat and intact lobule

areas) [8]. As well, other MRI studies found that adiposedegeneration or nodular parenchyma is considered the mostaccurate finding of SS patients’ salivary glands [10,18]. There-

fore, in our study we performed MRI using T1WI and STIRsequences to monitor the fat deposition in SS which mightbe useful in diagnosing and staging of the disease and assessing

its progress in patients whose clinical and serologic findings aresuggestive of the disease.

Regarding the gland size, the present study revealed parotid

enlargement in 46.7% patients. Other studies reported thatparotid enlargement may be a feature of SS in approximately50% of patients [27,28]. Glandular contour was irregular in46.7% patients. This finding was also reported in an ultrasono-

graphic texture analysis of the parotid gland in SS [29]. Lymphnode enlargement was observed in 73.3% patients. Parotidlymphoma was reported to occur in SS [30]; however, in this

study no evidence of parotid tumors was observed.In the present study, on correlating the ductal stages and SS

stages, no statistically significant correlation was found in both

CS and MRS. Also, on correlating the SS stages in MRS andMRI parenchymal stage, no statistically significant correlationwas found. On the other hand, a statistically significant posi-tive correlation was found between the MRI parenchymal

stage and the ductal stage of the primary and secondarybranches in MRS, this means that increase in destruction ofprimary and secondary branches is associated with an increase

in severity of parenchymal damage. These findings in line withthose who stated that, with disease progression, completedestruction of the minor duct radicals occur. Additionally,

once fat deposition develops the parenchyma will be lost for-ever [18,26,31].

On correlating the SS stage in both the CS and MRS with

the clinical and laboratory findings, no statistically significantcorrelation was found. On the other hand, a statistically signif-icant positive correlation was found between MRI parenchy-mal pattern and age, disease duration, anti Ro/La antibodies

and recurrent parotid swelling. This means that, an increasein age, disease duration, recurrent parotid swelling and antiRo/La were associated with an increase in heterogeneity of

parenchymal pattern. As in our study, Niemela et al. [18] sta-ted that the presence of anti-Ro/La antibodies was associatedwith the parenchymal findings on parotid MRI.

According to our results, the positive correlations found be-tween MRI stages and the clinical findings suggest that theMRI parenchymal staging used in our study may be used as

a diagnostic method to reflect the severity of the disease. More-over, the absence of correlation between the CS and MRS dis-ease staging and the clinical findings emphasizes theimportance of sialographic examination in the diagnosis and

staging of the disease.When sialectasis was taken as a reference for the diagnosis

of SS, MRS showed higher sensitivity and diagnostic accuracy

than CS (86.7% versus 78.6% respectively) which in closeproximity to Markusse et al. [32] who found that the diagnos-tic sensitivity of CS was 79%. On the other hand, other study

showed MRS sensitivity of 73% and diagnostic accuracy of83% [7]. Yu et al. [15] showed higher sensitivity and diagnosticaccuracy values (92% and 87.5% respectively) of MRS than

ours. In our study, one patient showed severe main duct dila-tation and irregularity without evidence of sialectasis in bothmodalities. Accordingly, this patient was falsely judged as neg-ative for SS. Therefore, we recommend considering the duct

abnormalities together with the presence of sialectasis in thediagnosis of SS. This increased the sensitivity and diagnosticaccuracy values of both MRS and CS (93.3% and 92.7%

respectively). The MRI parenchymal stage in our study dem-onstrated 93.3% sensitivity and diagnostic accuracy values.It is worth mentioning that one of our patients showed normal

sialographic findings, however, stage-I heterogeneous paren-chymal pattern was observed. Therefore, we attempted to com-bine MRS with MRI parenchymal pattern abnormalities for

further approach to a more accurate diagnosis. Interestingly,this combination showed 100% sensitivity and 100% diagnos-tic accuracy in the diagnosis of and staging of SS. Likewise, ina previous study a combination of T1WI and fat suppressed

T2WI MRI and MRS yielded a high sensitivity value (96%)[8].

Results of Kappa statistic for the inter-observer agreement

in interpreting the CS revealed moderate to good agreementregarding the main duct findings and the secondary ductalbranches. However, poor to fair agreement was found regard-

ing the remaining other findings. However, as regard MRS,there was a perfect agreement between all observers for themain duct; moderate to very good agreement for the sialecta-sis; good agreement was found between two observers for

the overall branching pattern and moderate agreement forthe ductal system branches. Also, there was a perfect agree-ment between all observers as regards parenchymal pattern

and gland contour in MRI. This denotes that the diagnosticreliability of MRS is higher than CS, thus MRS can securelyreplace CS in the diagnosis of SS.

In conclusion, from this preliminary study, MRI/MRS isvaluable in the evaluation of the salivary glands in SS sincethe clinical profile of the patients is not correlated with the

severity of the disease. Also, MRS is proved to be a sensitiveand highly accurate imaging for the diagnosis and staging ofSS and may securely replace CS in the evaluation of SS.MRI and MRS give information on different aspects of

154 N.S. Ahmed et al.

glandular pathology; therefore, both should be performedwhen examining the parotid glands in SS.

References

[1] Ramos-Casals M, Tzioufas AG, Font J. Primary Sjogren’s

syndrome: new clinical and therapeutic concepts. Ann Rheum

Dis 2005;64:347–54.

[2] Margaix-Munoz M, Bagan JV, Poveda R, Jimenez Y, Sarrion G.

Sjogren’s syndrome of the oral cavity. Review and update. Med

Oral Patol Oral Cir Bucal 2009;14(7):E325–30.

[3] Vitali C, Bombardieri S, Jonsson R, Moutsopoulos HM, Alex-

ander EL, Carsons SE, et al.. Classification criteria for Sjogren’s

syndrome: a revised version of the European criteria proposed by

the American–European Consensus Group. Ann Rheum Dis

2002;61(6):554–8.

[4] Al-Hashimi I. Xerostomia secondary to Sjogren’s syndrome in the

elderly: Recognition and management. Drugs Aging

2005;22(11):887–99.

[5] Salaffi F, Argalia G, Carotti M, Giannini FB, Palombi C. Salivary

gland ultrasonography in the evaluation of primary Sjogren’s

syndrome. Comparison with minor salivary gland biopsy. J

Rheumatol 2000;27:1229–36.

[6] Kalk WW, Vissink A, Spijkervet FK, Bootsma H, Kallenberg

CG, Roodenburg JL. Parotid sialography for diagnosing Sjo-

gren’s syndrome. Oral Surg Oral Med Oral Pathol Oral Radiol

Endod 2002;94:131–7.

[7] Tonami H, Higashi K, MatobaM, Yokota H, Yamamoto I, Sugai

S. A comparative study between MR sialography and salivary

gland scintigraphy in the diagnosis of Sjogren’s syndrome. J

Comput Assist Tomogr 2001;25(2):262–8.

[8] Takagi Y, Sumi M, Sumi T, Ichikawa Y, Nakamura T. MR

microscopy of the parotid glands in patients with Sjogren’s

syndrome: quantitative MR diagnostic criteria. AJNR

2005;26:1207–14.

[9] El-Miedany YM, Ahmed I, Mourad HG, Mehanna AN, Aty SA,

Gamal HM, et al.. Quantitative ultrasonography and magnetic

resonance imaging of the parotid gland: can they replace the

histopathologic studies in patients with Sjogren’s syndrome? Joint

Bone Spine 2004;71(1):29–38.

[10] Niemela RK, Takalo R, Paakko E, Suramo I, Paivansalo M, Salo

T, et al.. Ultrasonography of salivary glands in primary Sjogren’s

syndrome. A comparison with magnetic resonance imaging and

magnetic resonance sialography of parotid glands. Rheumatology

2004;43:875–9.

[11] Dugonjic S, Ajdinovic B, Stefanovic D, Jaukovic L. Diagnostic

validity of dynamic salivary gland scintigraphy with ascorbic acid

stimulation in patients with Sjogren’s syndrome: comparison with

unstimulated whole sialometry. Vojnosanit Pregl 2008;65(1):41–6.

[12] Tanaka T, Ono K, Ansai T, Yoshioka I, Habu M, Tomoyose T,

et al.. Dynamic magnetic resonance sialography for patients with

xerostomia. Oral Surg Oral Med Oral Pathol Oral Radiol Endod

2008;106:115–23.

[13] Lomas DJ, Carrol NR, Johnson G, Antoun NM, Freer CEL. MR

sialography. Radiology 1996;200:129–33.

[14] Gadodia A, Bhalla AS, Sharma R, Thakar A, Parshad R. MR

sialography of iatrogenic sialocele: comparison with conventional

sialography. Dentomaxillofacial Radiol 2011;40:147–53.

[15] Yu CQ, Zheng LY, Zhang J. The value of MRS in diagnosis of

Sjogren’s syndrome in parotid gland. Shanghai Kou Qiang Yi Xue

2007;16(3):247–50.

[16] Stiller M, Golder W, Eva Doring, Kleim K. Diagnostic value of

sialography with both the conventional and digital subtraction

techniques in children with primary and secondary Sjogren’s

syndrome. Oral Surg Oral Med Oral Pathol Oral Radiol Endod

1999;88:620–7.

[17] Rubin P, Holt JF. Secretary sialography in disease of the major

salivary glands. AJR 1957;77:575–98.

[18] Niemela RK, Takalo R, Paakko E, Suramo I, Takalo R, Hakala

M. Magnetic resonance imaging and magnetic resonance sialog-

raphy of parotid glands in primary Sjogren’s syndrome. Arth Care

Res 2001;45:512–8.

[19] Habermann CR, Graessner J, Cramer MC, Aldefeld D, Reitmeier

F, Weiss F, et al.. MR-sialography: optimization and evaluation

of an ultra-fast sequence in parallel acquisition technique and

different functional conditions of salivary glands. Rofo

2005;177(4):543–9.

[20] Petridis C, Ries T, Cramer MC, Graessner J, Petersen KU,

Reitmeier F, et al.. MR sialography: evaluation of an ultra-fast

sequence in consideration of a parallel acquisition technique and

different functional conditions in patients with salivary gland

diseases. Rofo 2007;179(2):153–8.

[21] Kalinowski M, Heverhagen JT, Rehberg E, Klose KJ, Wagner

HJ. Comparative study of MR sialography and digital subtraction

sialography for benign salivary gland disorders. Am J Neurora-

diol 2002;23:1485–92.

[22] Mafee MF, Valvassori GE, Becker M. Imaging of the head and

neck. 2nd ed. Germany: George Theme, Co; 2004, p. 625–81.

[23] Tonami H, Ogawa Y, Matoba M, Kuginuki Y, Yokota H,

Higashi K, et al.. MR sialography in patients with Sjogren’s

syndrome. AJNR 1998;19:1199–203.

[24] Ohbayashi N, Yamada I, Yoshino N, Sasaki T. Sjogren’s

syndrome: comparison of assessments with MR sialography and

conventional sialography. Radiology 1998;209:683–8.

[25] Izumi M, Eguchi K, Ohki M, Uetani M, Nagataki S, Kita M,

et al.. MR imaging of the parotid gland in Sjogren’s syndrome: a

proposal of new diagnostic criteria. AJR 1996;166:1483–7.

[26] Izumi M, Eguchi K, Nakamura H, Nagataki S, Nakamura T.

Premature fat deposition in the salivary glands associated with

Sjogren syndrome: MR and CT evidence. AJNR 1997;18(5):951–8.

[27] Sapp JP, Eversole LR, Wysocki GP. Contemporary oral and

maxillofacial pathology. 2nd ed. China: Mosby, co; 2004, p. 344–46.

[28] Saraf S. Textbook of oral pathology. 1st ed. India: Jaypee, Co;

2006, p. 372–3.

[29] Arjii Y, Ohki M, Eguchi K, Izumi M, Arjii E, Mizokami A, et al..

Texture analysis of sonographic features of the parotid gland in

Sjogren’s syndrome. AJR 1996;166:935–41.

[30] Theander E, Henriksson G, Ljungberg O, Mandl T, Manthorpe

R, Jacobsson L. Lymphoma and other malignancies in primary

Sjogren’s syndrome: a cohort study on cancer incidence and

lymphoma predictors. Ann Rheum Dis 2006;65:796–803.

[31] Bailoor DN, Nagesh KS. Fundamentals of oral medicine &

radiology. 1st ed. India, New Delhi: Jaypee; 2005, p. 206–19.

[32] Markusse HM, van Putten WIJ, Breedveld FC, Oudkerk M.

Digital subtraction sialography of the parotid glands in primary

Sjogren’s syndrome. J Rheumatol 1993;20:279–83.