Diagnostic utility of salivary gland ultrasonography in suspected primary Sjögren's disease: a comparison of OMERACT-based ordinal and summative scoring.

BACKGROUND

Salivary gland ultrasonography (SGUS) is a promising, non-invasive tool in the diagnostic workup of suspected primary Sjögren’s disease (SjD). This study aimed to compare the diagnostic accuracy of two SGUS scoring methods - an ordinal score (0–6) and a sum score (0–12) - derived from the OMERACT 0–3 grading system, using labial salivary gland biopsy as the reference standard.

METHODS

Sixty consecutive patients with suspected primary SjD underwent SGUS of the four major salivary glands. Each gland was graded using the OMERACT 0–3 system. Two composite scores per patient were calculated: (1) ordinal score (maximum 6), considering the two most affected glands and (2) sum score (maximum 12), obtained by adding the scores of all four glands. Diagnostic accuracy was assessed using ROC curve analysis, with histopathology as reference.

RESULTS

Labial biopsy was positive in 23 patients (38.3%). The ordinal score yielded an AUC of 0.687 (95% CI: 0.546–0.828), with 91.9% specificity (95% CI: 83.1–100%) at a threshold ≥ 4. The sum score showed an AUC of 0.683 (95% CI: 0.543–0.824), with 89.2% specificity (95% CI: 79.2–99.2%) at ≥ 5. SGUS scores ≥ 2 (ordinal) and ≥ 3 (sum) were significantly associated with anti-Ro52 antibody positivity (p = 0.017 and p = 0.007, respectively). At these high-specificity thresholds, labial biopsy could have potentially been deferred in 16.6% (ordinal ≥ 4) and 21.7% (sum score ≥ 5) of biopsy-positive cases, respectively.

CONCLUSIONS

Although overall diagnostic performance was modest, SGUS scoring based on OMERACT definitions demonstrated high specificity at selected thresholds. These findings support its potential role as a biopsy-sparing tool in selected patients with high pre-test probability and compatible serological profiles. Prospective studies are needed to validate these results and optimize threshold selection. TRIAL REGISTRATION: not applicable.