Evaluation of primary acquired nasolacrimal duct obstruction: Comparison of CT-DCG and dacryoendoscopy in accurately localizing the lacrimal drainage obstructions

Objective To correlate and evaluate the power and limitations of CT-DCG in determining the level and type of lacrimal duct obstruction in comparison to dacryoendoscopy in patients clinically suspected to be having partial or complete primary acquired nasolacrimal duct obstruction (PANDO).Methods A retrospective chart review was performed on 1232 lacrimal drainage systems of 957 patients who suffered from primary acquired nasolacrimal duct obstruction (PANDO) at Shanghai Ninth People's Hospital. Patients were examined with CT-DCG and correlated with dacryoendoscopy and the findings of clinical examination.Results Of the studied patients, 173 were men and 784 were women with an age range of 18-93 years. Of the 1232 lacrimal pathways, good CT-DCG images could be obtained in 980 cases and dacryoendoscopy in 957 cases. Of these complete obstructions were noted in 81% (794/980), and partial obstructions were identified in 19% (186/980) with CT-DCG. CT-DCG and dacryoendoscopy showed 68.4% agreement for the type of the obstruction and 63% for the level of the obstruction. The majority of the obstructions occurred at the sac-duct junction (62.5%) followed by the upper half of the nasolacrimal duct (27.5%). There was a significant difference in the correlation of the obstruction type with age group and with the duration of symptoms. As the duration of symptoms increased, the proportion of complete lacrimal duct obstructions as shown on CT-DCG images increased and the proportion of incomplete obstruction decreased (p = 0.015).Conclusions The junction of lacrimal sac and nasolacrimal duct was the most common obstruction site. Age and the duration of symptoms influenced the type of obstruction noted. The degree and level of agreement between the investigations was moderate. A combination of CT-DCG and Dacryoendoscopy could together identify the location more accurately.