OBJECTIVES: The minor salivary gland biopsy (MSGB) is widely considered an important component of the diagnostic algorithm of primary Sjögren's syndrome (pSS) and is mentioned in all the classification criteria sets for the disease. The aim of this study, coordinated by the Italian Society of Rheumatology, was to verify the inter-observer agreement on the evaluation of MSGB among different experienced Italian rheumatologic centres, in order to better standardise the diagnostic methodology. METHODS: Seven centres participated in the study, providing a total of 50 MSGB samples. Each center blindly classified all the samples according to the Chisholm and Mason (CM) grading. The results were collected and analysed. RESULTS: The inter-observer agreement was satisfactory when the samples were stratified as consistent and non-consistent with the final diagnosis of pSS (median κ =0.75; mean κ =0.70). Nonetheless, significant discrepancies in the histopathologic evaluation of MSGB emerged when the agreement was assessed on the single scores. Considering the modal CM grading for each sample as the correct grading, upon re-examination, a potential bias in the final clinical diagnosis was detected in 7 out of 50 samples. CONCLUSIONS: This study has shown significant discrepancies in the evaluation of MSGB among different rheumatologic centres in the same country. Greater standardisation of the procedure is clearly necessary, both to improve the diagnostic performance and scientific communication.
OBJECTIVES: The minor salivary gland biopsy (MSGB) is widely considered an important component of the diagnostic algorithm of primary Sjögren's syndrome (pSS) and is mentioned in all the classification criteria sets for the disease. The aim of this study, coordinated by the Italian Society of Rheumatology, was to verify the inter-observer agreement on the evaluation of MSGB among different experienced Italian rheumatologic centres, in order to better standardise the diagnostic methodology. METHODS: Seven centres participated in the study, providing a total of 50 MSGB samples. Each center blindly classified all the samples according to the Chisholm and Mason (CM) grading. The results were collected and analysed. RESULTS: The inter-observer agreement was satisfactory when the samples were stratified as consistent and non-consistent with the final diagnosis of pSS (median κ =0.75; mean κ =0.70). Nonetheless, significant discrepancies in the histopathologic evaluation of MSGB emerged when the agreement was assessed on the single scores. Considering the modal CM grading for each sample as the correct grading, upon re-examination, a potential bias in the final clinical diagnosis was detected in 7 out of 50 samples. CONCLUSIONS: This study has shown significant discrepancies in the evaluation of MSGB among different rheumatologic centres in the same country. Greater standardisation of the procedure is clearly necessary, both to improve the diagnostic performance and scientific communication.