OP0095 A DECISION MODEL OF LABIAL GLAND BIOPSY BASED ON B-MODE ULTRASONOGRAPHY WITH SHEAR-WAVE ELASTOGRAPHY IN PATIENTS WITH SUSPECTED SJÖGREN’S SYNDROME

Abstract

Background:Focal lymphocytic sialadenitis defined as focus score (FS) ≥1 on labial gland (LG) biopsy plays an integral role in various classification criteria of Sjögren’s syndrome (SS). However, suspected patients often hesitate to receive a biopsy; and rheumatologists hope a decision for biopsy based on a high predicted incidence of FS≥1, or against biopsy based on an absolutely low predicted incidence.Objectives:To build a decision model of LG biopsy based on B-mode ultrasonography (US) with shear-wave elastography (SWE) in patients with suspected SS.Methods:Patients who had at least one symptom of oral dryness (based on AECG questions) or had anti-SSA positive were recruited and signed a written informed consent. Bilateral parotid (PG) and submandibular glands (SMG) were examined with B-mode US which graded the echostructure of each gland on a scoring system scaled 0 to 4 (US score), and SWE which described the elasticity of glands. Then LG biopsy was performed.Results:(1)Ninety-one patients whose mean age was 43±15 years were enrolled and 93% of them were female. Anti-SSA was detected in 77 patients (85%) and 28 patients (31%) showed unstimulated whole saliva flow rate (USFR)≤0.1mL/mim. There were 57 patients (63%) showing FS≥1 on LG biopsy. Sixty-three patients (69%) were classified as primary SS, 10 patients (10%) were secondary SS, 18 patients (20%) were uCTD and one patient was RA without SS.(2)US scores were equal between PG and SMG in 59 patients (65%), while the rest patients showed different US scores between two glands: 7 patients (8%) showed higher US scores in PG and 25 patients (27%) showed higher scores in SMG. In each pair of glands US scores were equal. SWE values in PG or SMG of US score 1, 2 or 3 were significantly higher than those of US score 0, while SWE values in glands of US score 4 became declined and showed no significant difference from those with US score 0 (Figure 1A).(3)Heatmap showed US scores in either major salivary gland of patients with FS≥1 on LG biopsy were significantly higher than those with FS<1 (all p<0.001, Figure 1B). ROC curve showed a total US score (including bilateral PG and SMG) ≥9 and a total SWE value (including bilateral PG and SMG)≥30 could significantly recognize patients with FS≥1, respectively with specificity of 100% and 93% (Figure 1C). In this cohort, among 51 patients with a total US score ≥9 and/or a total SWE value≥30, 49 patients (96%) showed FS≥1 on LG biopsy; while two outliers showed total US scores were both 8 although combined SWE values≥30. Other 29 patients showed total US scores≤6 with total SWE values <30 and only one patient (3%) showed FS≥1 on LG biopsy. The remaining 11 patients showed total US scores were 8 with total SWE values <30 and 64% of them (n=7) showed FS≥1.Conclusion:A preliminary decision model of LG biopsy based on B-mode US with SWE in patients with suspected SS were built in Table 1. For example, rheumatologists should reassess the need for biopsy if the incidence of FS≥1 would be <5%. Another cohort of patients with suspected SS is needed for further validation.Table 1.A preliminary decision model of LG biopsy based on B-mode US with SWE in patients with suspected SSAlgorithm*Comments on the decision of LG biopsyA total US score≥9 and/or a total SWE≥30The specificity of FS≥1 on biopsy is >93%. Biopsy is recommended. In some special cases (e.g. contraindicated to biopsy), this item is a potential alternative to LG biopsy.A total US score 7~8 with a total SWE <30It is hard to predict the result of FS, so biopsy is strongly recommended.A total US score≤6 with a total SWE <30The incidence of FS≥1 would be <5%. Rheumatologists should reassess the need for biopsy.