MRI evaluation of lacrimal drainage after external and endonasal dacryocystorhinostomy.

PURPOSE External and endonasal dacryocystorhinostomy (EX-DCR and EN-DCR, respectively) affect the tear drainage mechanism. This study evaluates the preservation of "lacrimal pump" function in both procedures. METHODS Cases of successful EN-DCR (4 patients) and EX-DCR (4 patients) were included. All patients underwent MRI of the rhinostomy areas, at least 6 months postoperatively. The vertical diameter of rhinostomy (both osseous and soft-tissue apertures) was measured in T1-oriented images, whereas the signal intensity levels were examined for 3 regions of interest (ROIs) in T2-oriented (true fast imaging steady state pulse) images with instillation of normal saline to the conjunctival fornices, both before and after blinking (activation of the "lacrimal pump"). ROI 1 corresponded to the globe (control), ROI 2 corresponded to the inferior conjunctival fornix, and ROI 3 corresponded to the rhinostomy site. RESULTS Signal intensity in ROI 3 (rhinostomy) was significantly increased after blinking in both EX-DCR and EN-DCR cases. The increase was significantly higher in the latter. Signal intensity changes in ROI 3 were significantly correlated with rhinostomy size in both groups, whereas the respective correlations with the postoperative interval were not statistically significant. CONCLUSIONS Findings imply that the "lacrimal pump" is active following DCR and may be better preserved in the EN-DCR than in the EX-DCR group. Persistent epiphora after patent DCR may thus be attributed to a defective "pump" function and treated accordingly.