SAT0617 Salivary gland ultrasound can differentiate primary sjÖgren's syndrome from systemic diseases with salivary gland involvement

Abstract

Background: Ultrasound of the major salivary glands (SGUS) has shown high diagnostic accuracy for primary Sjogren's syndrome (pSS) [1,2]. None of the existing studies, however, included a control group of patients with sarcoidosis, amyloidosis, human immunodeficiency virus (HIV) or hepatitis C (HCV) infection [2]. These diseases are known to frequently affect the major salivary glands, cause dry mouth or may have histopathology similar to pSS. Objective(s): To assess the diagnostic accuracy of SGUS to differentiate pSS from systemic diseases with salivary gland involvement. Method(s): Twenty consecutive patients fulfilling the AECG criteria for pSS and 20 consecutive patients with well-established systemic diseases, i.e., 5 patients with sarcoidosis, 5 patients with amyloidosis, 5 patients with HIV infection and 5 patients with HCV infection, were included. Echogenicity, parenchymal homogeneity, presence of hypoechogenic areas and clearness of posterior glandular border of both parotid and submandibular salivary glands were scored independently by two blinded observers and rated according to the Hocevar et al scoring system (ultrasound total score (UTS), range: 0-48) [3]. The optimal cut-off point was determined by using ROC curve. The agreement between SGUS positivity and classification as pSS was assessed. Result(s): The median age of patients with pSS and systemic diseases was 50 and 53 years, respectively. Daily complaints of dry mouth longer than 3 months, need of liquid to swallow food and recurrent or persistent swelling of the major salivary glands were reported by 95%, 85% and 70% of patients with pSS compared to 45%, 30% and 20% of patients with systemic diseases. The accuracy of SGUS to discriminate pSS was excellent, with an area under ROC of 0.91 and optimal cut off point of 15. The agreement between SGUS and diagnosis for pSS was good (kappa=0.75, percentage of absolute agreement 87.5) with a sensitivity of 90%, specificity of 85%, positive predictive value of 86%, negative predictive value of 89% and diagnostic odds ratio of 51. UTS was positive in 2 patients with HIV infection and one patient with sarcoidosis. Patients with pSS had significantly higher UTS than patients with systemic diseases (median UTS 27 vs . 10, palpha0.001) and patients of the various subgroups (palpha0.05; Fig). Conclusion(s): This pilot study indicates that SGUS has a high diagnostic accuracy to discriminate pSS from associated systemic diseases with salivary gland involvement. A minority of HIV and sarcoidosis patients, however, may show SGUS findings mimicking pSS.