Value of ultrasound biomicroscopy in assessment of small masses at medial canthal region.

PURPOSE: Conventional imaging techniques are not sensitive enough to reveal detailed structures of lacrimal drainage system (LDS) and its surrounding tissue (ST). Our study aimed to explore utility of ultrasound biomicroscopy (UBM) in assessment of small masses at the medial canthal region and compare performance of UBM with conventional imaging techniques. METHODS: We prospectively recruited cases with small mass (long axis < 1 cm) at the medial canthal region (upper LDS-located area) from June 2017 to October 2018. UBM ± color Doppler flow imaging (CDFI) and conventional imaging techniques (computed tomography, magnetic resonance imaging, and dacryocystography) were conducted by four independent practitioners. Results were analyzed against gold standards with Cohen's kappa test in three aspects including LDS patency, mass location, and presumptive diagnosis. Corresponding gold standards were syringe and dacryocystography, intraoperative findings, and pathological/empirical diagnosis. RESULTS: Seventy-two cases were recruited, including 20 cases of LDS lesions and 52 cases of ST lesions. Female (odds ratio 7.14) and age ≥ 37 (odds ratio 9.80) were risk factors for LDS lesion, and age range of 15-25 (odds ratio 9.17) was a risk factor for inflammatory ST lesion. In terms of LDS patency, UBM results were reliable for the detection of pre-saccal obstruction (kappa = 0.920), but were not reliable for intra-saccal and post-saccal obstruction (kappa = 0.106). In terms of mass location, the UBM (kappa = 0.766) performed better than conventional techniques (except for dacryocystography) to sort out ST lesions, with sensitivity of 93.8% and specificity of 83.3%. In terms of diagnosis, the UBM (kappa = 0.882) outweighed conventional techniques (except for magnetic resonance imaging) to distinguish cysts from nodules, with sensitivity of 93.8% and specificity of 94.4%. Notably, the UBM + CDFI achieved better performance than the UBM when screen out inflammatory lesions (kappa = 0.926 vs kappa = 0.689) and LDS-adjacent lesions (kappa = 0.815 vs kappa = 0.673), resulting in sensitivity of 91.7% and specificity of 100% for both testing items. If deep lesions (at the lacrimal sac-harbored area) were excluded, UBM reliability to detect inflammatory lesions (kappa = 0.915) and LDS-adjacent lesions (kappa = 0.770) improved, achieving sensitivity of 90.0% and 88.9%, and specificity of 100.0% and 92.7%, respectively. CONCLUSIONS: The UBM is a valuable tool to assess superficial masses at the medial canthal region regarding pre-saccal obstruction, mass location, and presumptive diagnosis. TRIAL REGISTRATION: This work was registered on Chinese Clinical Trial Registry website with registration number ChiCTR1800018956 .