Background/Purpose: To determine whether hypoechoic lesions on parotid gland ultrasound can serve as
a surrogate marker for a focus score ≥1 on minor salivary gland biopsy in
patients with suspected Sjogren’s syndrome (SS).

Methods: We analyzed 220 parotid gland salivary gland
ultrasounds that were obtained on 220 patients being evaluated for possible or
known SS at the Johns Hopkins Sjogren’s Center
between 2012 and 2015. All scans were performed and interpreted by a
radiologist (JF) with expertise in ultrasonography.  The presence of hypoechoic lesions within the parotid gland parenchyma was
determined in each scan. Sjogren’s syndrome (SS) was
diagnosed according to AECG or preliminary ACR criteria. Statistical analyses
were performed using Fisher’s exact and Student’s t-test.

Results: Among the 220 patients, 116 fulfilled
classification criteria for SS. The remaining patients were labeled as “sicca controls”. Hypoechoic lesions
were significantly more frequent in SS patients versus controls (55/116 versus
3/104; p <0.0001) resulting in a sensitivity of 47% and specificity of 97%
for SS. A minor salivary gland biopsy was performed in 64 patients with SS and
showed a focus score ≥ 1 in 51 (80%). The presence of hypoechoic
lesions on ultrasound was highly associated with FLS ≥1 (23/24 versus
28/40, OR 9.9; 95% CI 1.2-81.5). The mean focus score was not significantly
different in SS patients with hypoechoic lesions compared
to SS patients without hypoechoic lesions (3.22
versus 2.39, p=ns). Positive
autoantibody status (anti-SSA/SSB/rheumatoid factor, hypergammaglobulinemia)
as well as leukopenia, monoclonal gammopathy and
salivary gland swelling were significantly more frequent in SS patients with hypoechoic lesions, whereas there was no difference in
frequency of C3 and C4 hypocomplementemia or symptom
duration. Other features on parotid gland ultrasound, including heterogeneity,
calcifications and presence of cysts did not predict FLS ≥ 1 as
accurately as hypoechoic lesions. Moreover, addition
of submandibular gland ultrasound data did not add additional predictive value.

Conclusion: Hypoechoic
lesions on parotid gland ultrasound strongly correlate with the presence of a
focus score ≥1 on minor salivary gland biopsy. Larger studies are needed
to determine whether ultrasound imaging may replace a diagnostic lip biopsy in
subsets of patients evaluated for SS.