Clinical outcomes and prognostic factors in patients with nasolacrimal duct obstruction or stenosis using dacryoendoscopy

Introduction

Lacrimal duct obstruction can manifest as epiphora or recurrent infections with mucopurulent discharge and may even cause social embarrassment. Obstruction can occur at any level along the lacrimal drainage system: punctum, canaliculus, nasolacrimal duct or nasal ostium.1 The aetiopathogenesis of lacrimal duct obstruction appears to be multifactorial.2 Dacryocystorhinostomy (DCR) is the preferred surgery for nasolacrimal duct obstruction (NLDO) but may be complicated by nasal bleeding, skin scar and granuloma.3

With the development of dacryoendoscope, silicone intubation has emerged as a common option to address NLDO, especially in China.4 5 Dacryoendoscope has been proven to be a useful instrument not only for understanding differences in the aetiology of obstructions but also for treating lacrimal duct obstructions.6 Furthermore, additional techniques, such as microdrill dacryoplasty, laser dacryoplasty, can achieve more effective tissue dissection than direct probing techniques.7 Although dacryoendoscope has increased the success rate of silicone intubation, the success rate ranges from 51% to 90%.8 The difference in success rates seems to depend on the site of obstruction, differences in the surgical technique, stringency of definitions of success, duration of follow-up and surgeon experience.5 9 10 To our knowledge, no study has comprehensively assessed the risk factors of treatment failure after silicone intubation.

A greater understanding of factors associated with treatment failure may allow us to better select suitable treatment procedures and ultimately improve treatment outcomes. The aim of our study was to evaluate the results of dacryoendoscopic-assisted laser dacryoplasty with silicone intubation (DLDI) for the management of NLDO or stenosis, and to identify factors associated with DLDI failure.

Methods

Surgical procedure

All surgeries were performed under local anaesthesia by one surgeon (HL). The infratrochlear nerve and infraorbital nerve were blocked with 2% lidocaine, and cotton gauze strips soaked in 4% lidocaine with epinephrine 1:100 000 solution were inserted into the inferior nasal meatus. The dacryoendoscope was inserted into the lacrimal duct through the upper lacrimal punctum. When the dacryoendoscope reached the site of obstruction (figure 1C), laser dacryoplasty (FOX diode laser, A.R.C. Laser, Nürnberg, Germany) was performed by widening the stenotic or obstructed section under dacryoendoscopic guidance without blind manipulation (figure 1D and figure 1E). The settings on the semiconductor diode laser were 4.5 watt and the laser energy was delivered by a 320 nm semirigid silica-polyamide laser fibre (figure 1B) using the contact mode. After removing the dacryoendoscope, a silicone tube (figure 1A) was inserted into the lacrimal passage. Double bicanalicular silicone intubation was performed as previously described.13 Correct insertion of the silicone tube was verified with the nasal endoscope (figure 1F) and anterior segment slit-lamp image showing the position of the silicone tube (figure 1G).

Request permission

Figure 1

Schematic view of DLDI for a patient with nasolacrimal duct obstruction. (A) Image depicting the soft probe and the silicone tubes. (B) The laser fibre (arrow) arranged in the laser inlets of dacryoendoscope. Intraoperative dacryoendoscopic views of laser dacryoplasty. Dacryoendoscope equipped with laser fibre (C, arrow) reached the site of obstruction (C, black asterisk) and then laser dacryoplasty was performed to widen the blocked section (D, E) until the opening of the nasolacrimal duct (E, black asterisk) was reached. (F) The view through the nasal endoscope (white asterisk) and dacryoendoscope (black asterisk) at the inferior meatus, soft probe (arrow) is seen passing through the opening of the nasolacrimal duct (white asterisk). (G) Anterior segment slit-lamp images showing the position of silicone tube (arrow) after DLDI. DLDI, dacryoendoscopic-assisted laser dacryoplasty with silicone intubation.

Results

Dacryoendoscopic examinations were performed on 1986 patients during the study period. The classification of the dacryoedoscopic findings is demonstrated in online supplemental figure 1. Online supplemental figure 2 shows preoperative and postoperative dacryoendoscopic images of the ostium at 24 months after the removal of the silicone tube for three patients of each group (online supplemental figure 2). There were 371 male and 1615 female participants. The age range was 21–80 years, with a mean age of 57.4±10.9 years. Online supplemental table 1 shows the baseline characteristics of the patients.

A total of 522 eyes from 522 patients with NLDO or stenosis who underwent DLDI were included. Of the 344 cases of complete NLDO, 49% (n=170) of the cases had a preoperative diagnosis of proximal NLDO, and 51% (n=174) had distal NLDO. The remaining cases were nasolacrimal stenosis (n=178). There were no statistically significant differences in patient age, sex or severity of epiphora among the three groups. The mean (SD) age of the patients was 53 (15.3) years (range, 22–79 years), and 81% (n=423) were women. At the end of the trial in November 2023, a total of 50 of 522 participants (9.6%) were lost to follow-up, and 69.5% of the participants completed a final visit. No differences were observed across the three groups with respect to the retention of participants.

The duration of symptoms for Group A was significantly longer than that for Groups B and C. The mean duration from surgery to removal of silicone tubes was 3.0±0.6 months. The resulting Kaplan-Meier estimated cumulative anatomical or functional success rates of DLDI at the 2-year follow-up were significantly different among the three groups (figure 2). The cumulative anatomical or functional success rate for Groups B and C was higher than that for Group A. After adjusting for all other covariates, the Cox proportional hazards regression model revealed a higher risk of failed DLDI surgery in patients with a history of chronic dacryocystitis (HR=3.07; 95% CI, 1.91 to 4.94; p<0.001) and a duration of epiphora longer than 1 year (HR=2.67; 95% CI, 1.56 to 4.58; p<0.001). Age and sex were not associated with a risk of failed DLDI in the multivariable model (figure 3). Table 1 reports the patient demographics and postoperative outcomes in the three groups within 2 years of follow-up. Online supplemental table 2 reports the patient postoperative outcomes in three groups at different months.

Request permission

Request permission

A total of 97 patients (18.6%) continue to experience persistent epiphora after removal of the silicone stent in the three groups. Functional epiphora, defined as patients continuing to experience persistent epiphora after anatomically successful surgery, occurred in 5% (n=26) of patients. Anatomical epiphora, defined as impatency on lacrimal irrigation with symptomatic epiphora, recurred at a mean of 16.1 months after removal of the silicone stent. For the cases of failed DLDI surgery, 15% (n=15) of patients declined any further treatment. After a mean further follow-up of 6 months, 59% (n=62) of patients with anatomical impatency underwent DCR with a 93% success rate; of the remainder with functional epiphora, 8% (n=8) had an eyelid-tightening procedure with a 60% success rate and 18% (n=12) underwent reintubation with a silicone stent, with a success rate of 65%. Of the 522 eyes, 12 patients had intraoperative complications, including orbital haemorrhage and punctal damage, and 25 patients had postoperative complications, including laceration of the canaliculi (cheese wiring), granuloma formation, prolapsed tubes and punctal adherence (figure 4).

Request permission

Discussion

Knowledge concerning the morphological characteristics of lacrimal passage obstruction using dacryoendoscope is revolutionary. Dacryoendoscope allows clearer visualisation of the lumen of the entire lacrimal drainage system, leading to an understanding of specific details of lacrimal duct obstructions. Our study revealed that various types of lacrimal duct obstruction may be related to differences in etiologies, enabling the selection of preferred surgical procedures based on dacryoendoscopic findings.

Single silicone tube lacrimal duct system intubation was first introduced in 1968 by Keith.14 The operative concept was to widen the blocked section and aimed at reconstituting the natural lacrimal duct pathway. Previous studies have reported success rates for silicone intubation with concomitant dacryoendoscope of the nasolacrimal duct stenosis varying between 73.9% and 82.9% with 1-year postoperative follow-up.15 16 Our results showed cumulative anatomical and functional success rates of 88.5% and 83.0%, respectively, at the 2-year follow-up after removal of the silicone stent, which is above that observed in the previous studies. In our study, DLDI offers direct therapy to the obstructed site, leading to minimal collateral damage external to the target zone. Hence, it is likely to enable greater safer and more efficient removal of fibrous closure of the lumen. Meanwhile, DLDI can be performed under local anaesthesia as an outpatient procedure in the office, and it has advantages such as a short surgery time, less risk of bleeding during surgery and early rehabilitation. The findings of the present study also demonstrate that compared with patients with proximal NLDO, patients with distal NLDO or stenosis were significantly less likely to experience symptomatic recurrence after DLDI. These observations suggest that success rates seem to depend on the position and degree of obstruction or stenosis. In contrast, Kamao used a sheath-guided bicanalicular intubation approach for NLDO.5 They followed patients for up to 6 months and found an 83% success rate for distal NLDO and a 91% success rate for proximal NLDO.5 One explanation for this is that differences in surgical techniques and sample size may have introduced further variation in the results. The turning point of the DLDI learning curve commonly refers to the point at which a high surgical success rate is maintained and severe postoperative complications are reduced. After the turning point, the surgeon will have sufficiently mastered the technique. Our preliminary results showed a significant improvement in success rate in DLDI when comparing the first 45 cases with the second 45 cases illustrating a higher success rate in the second stages of training. (The relevant results are being summarised and will be reported in subsequent articles.)

Other risk factors associated with recurrence following initial DLDI surgery included a longer duration of epiphora and a history of chronic dacryocystitis. Lacrimal duct functional structures are important to prevent invasion of pathogenic agents and to regulate tear drainage.17 18 Previous studies have demonstrated that loss of functional structures, such as remodelling of structural epithelial, fibrosis of connective tissues and destruction of specialised blood vessels, which eventually leads to complete obstruction of the lacrimal duct and chronic dacryocystitis.19 The longer duration of epiphora means that there is a greater risk of pathogenic agent invasion, which gradually leads to loss of functional structures. Thus, patients with a longer duration of epiphora or a history of chronic dacryocystitis are more likely to have unusual clinicopathology that may put them at greater risk of recurrence.

Functional epiphora after DLDI among patients with NLDO or stenosis appears to be uncommon. According to the recommended treatment algorithm for functional epiphora, symptom improvement in most patients can be achieved with silicone reintubation and/or lower eyelid tightening.20 A bony obstruction of the nasolacrimal duct is not common and often occurs in patients with a history of bony congenital nasolacrimal duct obstruction. The disease is characterised by the hypoplasia of the middle or distal segment of the nasolacrimal canal evident through diagnostic imaging.21 Unlike the membranous type, bony nasolacrimal duct obstruction is not amendable to laser dacryoplasty. In our study, the exclusion criteria included congenital or traumatic lacrimal duct obstruction. For patients with a history of congenital nasolacrimal duct obstruction, we routinely performed dacryocystogram to rule out bony nasolacrimal duct obstruction.

The rate of adverse effects was not significantly different among the three groups, with the most common complications being cheese wiring and punctal granuloma formation. Multiple factors confer the mechanism of cheese wiring, including damage to the punctum and excessive bicanalicular tube loop tightening.22 There was a distinction of the dense connective tissue circularly surrounding the puncta. The tissue arrangement of puncta has a greater biomechanical effect to prevent wire cutting than the canaliculus, in which muscle fibres run parallel to it.23 Care should be taken to avoid injury of the punctum during the operation, thereby reducing the risk of cheese wiring. Additionally, administered antibiotic-corticosteroid eye drops may be helpful in reducing granuloma formation after surgery.24

The strengths of this study are its longitudinal analysis based on a large sample size. However, this study also had several potential limitations. First, the follow-up period of this study was relatively short, and we were unable to include patients from different ethnic groups. Second, the number of eyes was gradually lost to postoperative follow-up during the study period. It is difficult to maintain a patient’s follow-up for 24 months or longer, particularly for those with satisfactory surgery outcomes. Thus, increasing the follow-up time since the removal of the silicone stent may further decrease the success rate. Conducting studies with a patient’s follow-up for 5 years or longer in a multicentre study is required.

Conclusions

Our study revealed various types of lacrimal duct obstruction using dacryoendoscopy, enabling the selection of preferred surgical procedures for lacrimal duct obstruction. DLDI seems to yield a higher success rate for patients with distal NLDO or nasolacrimal duct stenosis, cause minimal tissue invasion and fewer significant complications. A higher risk of reoperation was associated with a longer duration of epiphora and a history of chronic dacryocystitis. These findings may allow us to better inform practice patterns when counselling patients about treatment options.