Cheng, Chiao-FengChiao-FengChengChen, Tseng-ChengTseng-ChengChenHsieh, Min-ShuMin-ShuHsiehHuang, Chung-YenChung-YenHuangKao, Jui-HungJui-HungKaoLan, Ting-YuanTing-YuanLanLin, Yu-HengYu-HengLinWang, ChiChiWangHuang, Hsien-NengHsien-NengHuangLin, Long-WeiLong-WeiLinChiang, Yu-JuYu-JuChiangCheng, Mei-FangMei-FangChengChu, Hsiao-SangHsiao-SangChuYI-MIN HUANGLu, Cheng-HsunCheng-HsunLuLi, Ko-JenKo-JenLiCHIEH-YU SHENHsieh, Song-ChouSong-ChouHsieh2025-12-312025-12-312025-12https://scholars.lib.ntu.edu.tw/handle/123456789/734931Ultrasound-guided core needle biopsy (CNB) of major salivary glands is a less invasive alternative to minor salivary gland biopsy for diagnosing Sjögren's disease (SjD), reducing risks of neurological deficits and pain. However, the optimal CNB specimen length for adequate glandular surface area remains uncertain. This study aimed to determine and validate the optimal CNB specimen length thresholds. This retrospective, dual-phase study included 119 consecutive patients undergoing submandibular gland CNB with an 18-gauge needle for suspicious chronic inflammatory sialadenitis. Specimen length, total surface area, and glandular surface area were recorded. A validation cohort (n=37) was analysed separately. Statistical analyses included correlation, regression, and ROC curve analysis. Specimen length correlated with glandular surface area (ρ=0.69, p<0.001). Multivariable analysis confirmed specimen length as a positive predictor (p<0.001) and fatty infiltration as a negative predictor (p=0.003) of glandular surface area. ROC analysis identified 7.6 mm as optimal for glandular surface area ≥4 mm², and 10.5 mm for ≥8 mm². Clinically significant haematomas occurred in 1.7% of cases. In the validation cohort, using the 7.6 mm threshold, PPV was 86.2%, and NPV 87.5% for glandular surface area ≥4 mm². For the 10.5 mm threshold, PPV was 76%, and NPV 100% for ≥8 mm². A CNB specimen length of ≥7.6 mm is sufficient for diagnosing SjD, while ≥10.5 mm may be required for clinical trials. These findings may support the integration of specimen length thresholds into CNB procedural guidelines.enjournal article10.55563/clinexprheumatol/0nn5rk41410586