Randomized Trial on Silicone Intubationin Endoscopic MechanicalDacryocystorhinostomy (SEND)for Primary Nasolacrimal Duct Obstruction

Kelvin K.L. Chong, MBChB(Hon), FCOphth,1,2 Frank H.P. Lai, MBChB, MRCSEd,1

Mary Ho, MBChB, MRCSEd,1 Abbie Luk, MBBS, MRCSEd,1 Ben W. Wong, FCOphth, FHKAM(Ophth),1

Alvin Young, FRCS, FCOphth1

Purpose: To study the effect of bicanalicular silicone intubation on endonasal endoscopic mechanicaldacryocystorhinostomy (EEM-DCR) for primary acquired nasolacrimal duct obstruction (PANDO).

Design: Randomized clinical trial.Participants: A total of 120 consecutive adults (103 females) with a presenting age of 64�13.7 years (range,

39e92 years) underwent EEM-DCR for PANDO from November 2005 to May 2009 in a lacrimal referral center.Methods: The EEM-DCR was performed by 2 lacrimal surgeons using standard techniques. Patients were

randomly assigned to receive or not receive bicanalicular silicone intubation for 8 weeks. No antimetabolite wasused. All patients received a course of oral antibiotics during nonabsorbable nasal packing for flaps tamponade,which was removed at the first postoperative visit. Patients were assessed at 1, 3, 6, 12, 26, and 52 weeks afterthe operation.

Main Outcome Measures: Surgical success was defined by symptomatic relief of epiphora, reestablishmentof nasolacrimal drainage confirmed by irrigation by 1 masked observer, and positive functional endoscopic dyetest by the operative surgeon at 12 months postoperatively. Intraoperative and postoperative complications wererecorded.

Results: A total of 118 of the 120 randomized cases completed 12 months of follow-up. Two patients died ofunrelated medical illnesses during follow-up. At 12 months postoperatively, there was no statistical difference inthe success rate between patients with (96.3%) and without (95.3%) intubation (P¼0.79). The odds ratio of failurewithout silicone intubation was 1.28 (95% confidence interval, 0.21e7.95). There was no difference in the inci-dence (P¼0.97) or the time to develop (P¼0.12) granulation tissue between the 2 groups. No significant differencewas found between successful and failed cases in terms of age (P¼0.21), sex (P¼0.37), laterality (P¼0.46), modeof anesthesia (P¼0.14), surgeon (P¼0.26), use of stent (P¼0.79), or presence of granulation tissue postoperatively(P¼0.39).

Conclusions: The current study design provided 90% statistical power to detect more than 21% differencein surgical outcome, and no such difference was found whether intubation was used or not used in EEM-DCR forPANDO at the 12-month follow-up.

Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussedin this article. Ophthalmology 2013;120:2139-2145 ª 2013 by the American Academy of Ophthalmology.

Treatment of primary acquired nasolacrimal duct obstruc-tion (PANDO) involves creating a fistulous bypass betweenthe lacrimal sac and the nasal cavity during dacryocysto-rhinostomy (DCR).1 As coined by Goldberg,2 the primaryaim of DCR is to create an adequate, atraumatic bony andsoft tissue ostium followed by controlled mucosal healing.Bicanalicular silicone intubation is thought to preventsealing of the edges of the lacrimal sac and impedefibrous closure during healing.3

There are a number of endonasal DCR studies per-formed without intubation, reporting success rates ranging

� 2013 by the American Academy of OphthalmologyPublished by Elsevier Inc.

from 87% to 93%,4e7 using endoscopic visualization ofa patent ostium as an end point. Recent reports by Unluet al,4 Madge and Selva,8 and Saiju et al9 also support thatcanalicular intubation does not play an important role inexternal or endonasal DCR for PANDO. However,there is a lack of prospective, randomized trials studyingthe efficacy of silicone intubation in PANDO.8,10

We hypothesized that bicanalicular silicone intubationdid not improve the surgical outcome of endonasal endo-scopic mechanical (EEM) DCR for PANDO at 12-monthfollow-up.

2139ISSN 0161-6420/13/$ - see front matterhttp://dx.doi.org/10.1016/j.ophtha.2013.02.036

Figure 1. Osteotomy was set superiorly at least 2 mm above the commoninternal punctum down to the sac-duct junction inferiorly in an opened sacwith fluorescein flushing through.

Figure 3. Left-sided, unopened agger nasi cell (*) and the operculum of themiddle turbinate being partially removed during endonasal endoscopicmechanical dacryocystorhinostomy (EEM-DCR) (endoscopic intraoperativeview).

Ophthalmology Volume 120, Number 10, October 2013

Methods

Consecutive patients with PANDO who were referred to thelacrimal clinic at the university hospital were recruited. The studywas conducted according to the Declaration of Helsinki, withapproval obtained from the clinical research ethics committee.Informed consent for operation and randomization were obtainedfrom participants. Both eyes of any eligible patient were allowed tobe recruited. Slit-lamp examination of tear film, eyelids, puncta,and anterior segments was performed to rule out reflex tearing, lidlaxity, and punctal stenosis. The presence and level of lacrimalobstruction were assessed by lacrimal irrigation and probing. Nasalendoscopy was performed to rule out significant nasal pathology.

Inclusion criteria were (1) adult (aged >18 years) with PANDOconfirmed intraoperatively after lacrimal sac exposure and (2)informed consent for the study and randomization.

Exclusion criteria were (1) canalicular obstruction requiringmembranectomy intraoperatively; (2) suspected lacrimal sac

Figure 2. Left-sided orbicularis muscle anterior to the unopened lacrimalsac, middle turbinate medially, sac fundus above, and sac-duct junctionbelow (endoscopic intraoperative view). Yellow solid line ¼ lacrimal sac(medial aspect); green dashed line ¼ orbicularis muscle (medial aspect).

2140

malignancy; (3) presence of lower lid or punctual ectropion,significant horizontal laxity, or facial nerve palsy; (4) previouslacrimal operation; and (5) previous irradiation, trauma, or majordiseases affecting the respective side of nose and orbit.

Standard EEM-DCR with opposing mucosal flaps11,12 was per-formed by 2 attending surgeons. Surgery was performed undergeneral or local anesthesia, according to the patient’s preference.Osteotomy was set superiorly at least 2 mm above the commoninternal punctum down to the sac-duct junction inferiorly (Fig 1).Anterior-superiorly, the orbicularis muscle was often exposed (Fig2), and the agger nasi cells or the operculum of the middleturbinate was frequently entered posterosuperiorly for exposure of

Figure 4. Left-sided, opened, andmarsupialized lacrimal sacwith anterior andposterior flaps (endoscopic intraoperative view; same patient as in Fig 2).Yellow solid line ¼ opened lacrimal sac (medial aspect); yellow dashed line¼ opened lacrimal sac (lateral aspect).

Figure 5. Before (A) and immediately after (B) left-sided endoscopic posterior bony septoplasty with improved visualization of middle turbinate.

Chong et al � Randomized Trial of Intubation in Endoscopic DCR

the lacrimal sac fundus13 (Fig 3). The lacrimal sac was then incisedand marsupialized with anterior and posterior flaps (Fig 4). Noantimetabolite was used. Five patients received concomitantendoscopic posterior bony septoplasties (Fig 5A and B) beforeEEM-DCR by the lacrimal surgeons to improve middle meatalaccess.

Randomization was carried out by a research assistant who wasnot involved in the clinical work of this study. Randomization tookplace after the lacrimal sac was opened intraoperatively to confirmthe absence of canalicular obstruction or lacrimal sac pathology.No bilateral simultaneous operation was performed, and each eyewas randomized separately for patients with bilateral PANDO. Aset of computer-generated random numbers were obtained beforerandomization and kept in a safety lock. Surgeons and patients hadno access to the random numbers throughout the study period. Therandom numbers were generated in blocks of 10, with equaldistribution between either group. Each case was randomlyassigned to receive or not receive bicanalicular silicone intubationfor 8 weeks.

Previous review at our lacrimal clinic on EEM-DCR revealedthat the success rate at 12 months postoperatively was approxi-mately 95% with silicone stent and 75% without silicone stent(Chong KKL, unpublished data, July, 2005). By using a 1:1 ratioof control to treatment eyes, a sample size of 65 eyes per group willgive a statistical power of 90% with a type 1 error of 0.05 by chi-square test to detect a difference of 20% in outcome. A total of 130eyes were thus recruited to this study.

All patients received a course of oral antibiotics (ampicillin/sulbactam [Unasyn; Pfizer Inc., New York, NY] 375 mg twice

Table 1. Baseline Characteristics (

With Silicone Tube (65 Cases)

Age, yrs (mean � SD [range]) 64.1�13.6 (39e87)Female:male 57:8 (87.7%:12.3%)Right:left 35:30 (53.8%:14.2%)General:local anesthesia 32:33 (49.2%:50.8%)Surgeon A:surgeon B 23:42 (35.4%:64.6%)Osteotomy size (mm) 18�2

SD ¼ standard deviation.yTwo-tailed independent t test.zTwo-tailed chi-square test.

daily) and nonabsorbable nasal packing for flaps tamponade, whichwas removed at the first postoperative visit. Patients were assessedat 1, 3, 6, 12, 26, and 52�2 weeks postoperatively.

Surgical success14,15 was defined as (1) symptomatic relief ofpreoperative epiphora or mucocele, (2) patency of tear drainage asconfirmed by lacrimal irrigation performed by 1 masked observer(an ophthalmology resident), and (3) spontaneous fluorescein stainflow into the rhinostomy after application to the conjunctivalcul-de-sac (functional endoscopic dye test [FEDT])16 by one of theoperative surgeons at 12 months postoperatively.

Anatomic failure was defined as regurgitation on irrigation anda closed intranasal ostium. Functional failure was defined as returnof epiphora and negative FEDT but patent on irrigation and anopened intranasal ostium.

Any intraoperative or postoperative complication was recorded,including the postoperative appearance of granulation tissue in theintranasal ostium. Endoscopic photographs of the intranasal ostiawere taken at each visit. Photographs 12 months postoperativelywere graded by one of the operating surgeons masked to the groupassignment. The ostium size was compared with standardizedphotographs of a large ostium and a small ostium.

Statistical analysis was performed with commercially availablesoftware: Microsoft Office Excel 2007 (Microsoft Corp, Redmond,WA) and SPSS 16 (SPSS Inc., Chicago, IL). Results of the first eyewere used for primary analysis. A separate set of data including the10 second eyes recruited per protocol were analyzed using a linearmixed model (Stata v. 10.0; StataCorp, College Station, TX) toaccount for potential correlation between eyes from the same indi-viduals. Numeric data were expressed as mean� standard deviation.

Second Eye Included, n¼128)

Without Silicone Tube (65 Cases) P Value

63.1�13.8 (39e92) 0.45y

56:9 (86.2%:13.8%) 0.80z

38:27 (58.5%:41.5%) 0.60z

30:35 (46.2%:53.8%) 0.60z

24:41 (36.9%:63.1%) 0.86z

16�3 0.23y

2141

Table 2. Baseline Characteristics (First Eye Only, n¼118)

With Silicone Tube (56 Cases) Without Silicone Tube (64 Cases) P Value

Age, yrs (mean � SD, range) 64.7�13.3 (39e87) 64.1�13.7 (39e92) 0.81y

Female: male 48:8 (85.7%:14.3%) 55:9 (85.9%:14.1%) 0.97z

Right: left 30:26 (53.6%:46.4%) 38:26 (43.8%:56.3%) 0.52z

General: local anesthesia 29:27 (51.8%:48.2%) 29:35 (45.3%:54.7%) 0.48z

Surgeon A: surgeon B 21:35 (37.5%:62.5%) 23:41 (35.9%:64.1%) 0.86z

Osteotomy size (mm) 17�2 16�3 0.41y

SD ¼ standard deviation.yTwo-tailed independent t test.zTwo-tailed chi-square test.

Ophthalmology Volume 120, Number 10, October 2013

Chi-square tests were used for comparison of categoric variables andproportions. Continuous variables were compared by unpairedt tests. AllP values were 2-tailed and considered significant if<0.05.

Results

A total of 171 lacrimal referrals were evaluated. After excluding51 patients because of canalicular obstruction (n¼30), revisionlacrimal surgeries (n¼18), history of irradiation (n¼2), andmidfacial trauma (n¼1), 120 patients with 130 cases of PANDO(10 bilateral cases) were randomized (Table 1). Two patients(1 female), aged 77 and 80 years, died of unrelated medicalillnesses during follow-up. Both patients were in the intubatedgroup with follow-up to 4 and 7 weeks postoperatively. The resultsof 118 patients (102 female) were available at the 12-monthfollow-up.

The mean age of patients was 64�13.7 years (range, 39e92years). A total of 103 patients (85.8%) were female. Sixty-fivepatients (50%) received no intubation. By using only the first eyefrom each recruited patient (n¼118), the intubated and controlgroups were comparable in terms of age, sex ratio, side of opera-tion, anesthesia method, surgeons, and osteotomy size (Table 2).

Among the 118 patients (first eye), 113 (95.8%) achievedsurgical success at the 12-month follow-up. Among the 5 failedcases, 2 patients had silicone intubation and 3 patients did not.Three patients had closed ostia (anatomic failure) and 2 patientshad open ostia and patent on lacrimal irrigation but return of epi-phora with negative FEDT (functional failure) (Table 3).

For the 2 nonintubated patients with anatomic failure, there wasprogressive mucosal fibrosis leading to closure of the ostia. Onepatient also developed mucosal adhesion of the ostium with middleturbinate (Fig 6A, available at http://aaojournal.org). Onenonintubated patient developed a return of epiphora and negativeFEDT 7 months after EEM-DCR, although syringing demon-strated anatomic patency (i.e., functional failure) (Fig 6B, availableat http://aaojournal.org).

One intubated patient had granuloma formation and closure ofthe ostium at 4 months after EEM-DCR, (Fig 6C, available athttp://aaojournal.org). Another intubated patient developedtearing sensation but with patency demonstrated on syringing

Table 3. Group Breakdown of Failed Cases at 12 MonthsPostoperatively

Intubated Nonintubated

Anatomic failure 1 2Functional failure 1 1

2142

(Fig 6D, available at http://aaojournal.org). Except for the lastpatient who refused further intervention, the other 4 patientsreceived revision EEM-DCR with intraoperative mitomycinC application and bicanalicular intubation and achieved anatomicand functional patency at their last follow-up visits.

By using only the first eye (n¼118) for analysis, there was nostatistical difference in surgical success between patients with(96.3%) and without (95.3%) intubation (P¼0.79) at 12 monthspostoperatively. The odds ratio of failure without silicone intuba-tion was 1.28 (95% CI, 0.21e7.95). With 54 patients in 1 groupand 64 patients in the other, the study had 90% statistical power todetect a 21% difference in surgical outcome between the groups,and no such difference was found. No significant difference wasfound in the mean time to failure (P¼0.14), ostial size at 12 months(P¼0.16), incidence (P¼0.97), or time to development (P¼0.12)of granulation tissue between the 2 groups (Table 4).

When second eyes also were included for analysis per protocol(n¼128), no statistical difference in surgical success was foundbetween patients with (96.8%) and without (95.3%) intubation(P¼0.67, linear mixed model) at 12 months postoperatively. Theodds ratio of failure without silicone intubation was 1.48 (95% CI,0.24e9.14). With 63 patients in the other and 64 patients in 1group, the study had 90% power to detect a 20% difference insurgical outcome between the groups, and no such difference wasfound. No significant difference was found in the mean time tofailure (P¼0.14), ostial size at 12 months (P¼0.24), incidence(P¼0.70), or time to development (P¼0.13) of granulation tissuebetween the 2 groups (Table 5).

At 12 months postoperatively, there was no significant differ-ence between successful and failed cases in age (P¼0.21), sex(P¼0.37), laterality (P¼0.46), surgeon (P¼0.26), use of stent(P¼0.79), presence of granulation (P¼0.39), or mode of anesthesia(P¼0.14) (Table 6).

A history or subsequent development of contralateral PANDO(i.e., bilaterality) was significantly more prevalent among failedcases (3 of the 5 failed cases were patients with bilateral PANDO)than successful cases (17 of 123 successful cases were patientswith bilateral PANDO) (P¼0.005) (Table 7).

Three patients (2.3%) had intraoperative orbital fat prolapsewhenthe lateral wall of the lacrimal sac was violated. Eight patients (6.3%)had postoperative nasal mucosal adhesion. Seven patients (5.5%)had epistaxis upon the removal of packing material at the firstpostoperative visit. In the intubated group (n¼63), 4 patients (6.3%)had tube prolapse, which was endoscopically reduced in the office inall patients. Three of these 4 patients had a successful outcome at 12months. One patient had return of epiphora with an open ostium(functional failure). One patient (1.6%) had stent extrusion 4 weekspostoperatively. This patient achieved surgical success at the 12-month follow-up. One patient (1.6%) had lacrimal false passageduring probing intraoperatively.

Table 4. Outcome at 12 Months Postoperatively (First Eye Only, n¼118)

With Silicone Tube (54 Cases)* Without Silicone Tube (64 Cases) P Value

Success at 12 mos (yes:no) 52:2 (96.3%:3.7%) 61:3 (95.3%:4.7%) 0.79y

Time to failure (mos) 4.0�1.4 7.0�1.7 0.14z

Presence of granulation tissue (yes:no) 23:31 (42.6%:57.4%) 27:37 (42.2%:57.8%) 0.97y

Time to granulation tissue development (wks) 5.9�3.4 4.3�3.8 0.12z

Ostial appearance at 12 mos (large:small) 12:42 (22.2%:77.8%) 8:56 (12.5%:87.5%) 0.16y

*Two patients died of unrelated medical illnesses.yTwo-tailed chi-square test.zTwo-tailed independent t test.

Chong et al � Randomized Trial of Intubation in Endoscopic DCR

Discussion

Although external DCR has been considered the goldstandard in lacrimal operation, the more recently publishedseries of endonasal DCR in both ophthalmic and oto-rhinolaryngologic literature report higher success ratescompared with prior studies.11,12 This likely reflects anincreased familiarity with endoscopic instrumentation andanatomy among lacrimal surgeons and an improved under-standing and control of nasolacrimal mucosal healing.2

Evidence continues to show that the endonasal approachoffers distinct advantages over external DCR. EndonasalDCR preserves lacrimal pump function without damagingthe deep pretarsal orbicularis muscle (Horner’s muscle) andthe peripheral fibers of the zygomatic and buccal branchesof the facial nerve.17 It also avoids disinsertion of the medialcanthal tendon.18 Full-thickness scarring from skin to nasalmucosa may occur after external DCR, which may form ringcontracture and negatively affect lacrimal sac movementduring blinking.19,20 Ancillary procedures, such as septo-plasty and middle turbinoplasty/turbinectomy, can be per-formed simultaneously to improve intraoperative access andto minimize postoperative adhesion.

Progressive ostial fibrosis with or without adhesion to thenasal septum or middle turbinate, inadequately sized orinappropriately located bony and soft tissue ostia, andunrecognized or postoperative canalicular obstruction arecommon causes of failure in DCR.21 On the other hand,mucosal healing inevitably reduces the size of internalostium after DCR.22,23 Various adjuvant treatments,including antimetabolites24 (i.e., 5-fluorouracil, mitomycin-

Table 5. Outcome at 12 Months Postoper

With Silicone Tube (6

Success at 12 mos (yes:no) 61:2 (96.8%:3.2Time to failure (mos) 4.0�1.4Presence of granulation tissue (yes:no) 25:38 (39.7%:60Time to granulation tissue development (wks) 5.4�1.9Ostial appearance (large:small) 12:51 (19.1%:80

*Two patients died of unrelated medical illnesses.yLinear mixed model.zTwo-tailed independent t test.

C, absorbable packing [e.g., Gelfoam {Pharmacia &Upjohn Co., New York, NY} with25 or without topicalsteroid11,12] or Merogel [Medtronic Xomed, Jacksonville,FL]26) have been used to maintain the patency of the ostium.Mechanical stents8 of different materials, with or withoutexpansile properties, have been used to prevent closure ofthe “soft tissue” ostium around the internal punctum.

The use of transcanalicular stent was first introduced byGraue and Glenie, who placed a silver wire through thelower punctum into the nose.8 Gibbs27 then describeda technique of silicone rubber tubing for intubation inDCR, although Quickert and Dryden28 were often creditedfor their contribution to this advance. In their reviews oflacrimal surgery, Anderson29 and Patrinely30 stated thatsilicone canalicular intubation was “probably the mostimportant recent advance in lacrimal surgery” but it was“not necessary for routine DCR.”

The added benefit of silicone intubation in PANDOduring EEM-DCR was not demonstrated by previousstudies.31 In a retrospective study, success was achieved in85.7% of the intubated cases versus 81.3% ofnonintubated cases4 at a mean follow-up of 23 monthsusing symptomatic improvement as the successful criterion.In a recent randomized trial of 46 cases of PANDO usingEEM-DCR, surgical success, defined by symptomatic reliefand patency to irrigation at 6 months postoperatively, wasachieved by 100% for the nonintubated cases versus 78%for the intubated cases.32 Unlu et al4 suggested that siliconeintubation may predispose to granulation formation withsubsequent rhinostomy closure. To date, there is a lack oflarge-scale, randomized trials with a sufficient follow-up

atively (Second Eye Included, n¼128)

3 Cases)* Without Silicone Tube (65 Cases) P Value

%) 62:3 (95.4%:4.6%) 0.68y

7.0�1.7 0.14z

.3%) 28:37 (43.1%:56.9%) 0.70y

6.6�3.1 0.13z

.9%) 8:57 (12.3%:87.7%) 0.24y

2143

Table 6. Comparison of Successful and Failed Cases at Postoperative 12 Months (First Eye Only, n¼118)

Success at 12 Mos (113 Cases) Failure at 12 Mos (5 Cases) P Value

Age, yrs 64.7�13.4 (39e92) 57.0�10.2 (43e68) 0.21*Female:male 97:16 (85.8%:14.2%) 5:0 (100%:0%) 0.37y

Right:left 64:49 (56.6%:43.4%) 2:3 (40%:60%) 0.46y

General anesthesia:local anesthesia 52:61 (46.0%:54.0%) 4:1 (80%:20%) 0.14y

Surgeon A:surgeon B 40:73 (35.4%:64.6%) 3:2 (60%:40%) 0.26y

Stent (yes:no) 52:61 (46.0%:54.0%) 2:3 (40%:60%) 0.79y

Granulation tissue (yes:no) 46:67 (41.2%:58.8%) 3:2 (60%:40%) 0.39y

*Two-tailed independent t test.yTwo-tailed chi-square test.

Ophthalmology Volume 120, Number 10, October 2013

period to study the efficacy of silicone intubation in EEM-DCR for PANDO.10,31

In this study, silicone intubation did not affect theoutcome (P¼0.79) at 12 months’ follow-up, the incidence(P¼0.97), and the time (P¼0.12) to develop granulationtissue. Success rates of the intubated and nonintubatedgroups using strict criteria16 were 96.3% and 95.3%,respectively. The results of our randomized study concurwith the existing literature. Surgical success ina nonrandomized series of 38 cases of PANDO wascomparable with (89.5%) and without (94.7%) intubationafter a median follow-up of 8 years.33 Unlu et al,4 Madgeand Selva,8 and Saiju et al9 supported that canalicularintubation did not play an important role in the outcomeof external or endonasal DCR performed for PANDO.Silicone intubation may lead to stent prolapse or loss,ocular surface irritation, laceration of canaliculi, andcreation of false passage during insertion.3,8,31 Moreover,retained stent material may cause failure. The results of ourstudy do not support routine intubation in EEM-DCR forPANDO. However, because our study was limited to casesof PANDO, the benefit of bicanalicular silicone intubationin other disorders, such as concomitant canalicularobstruction, needs to be further investigated.

In a histologic study comparing lacrimal sac biopsiesobtained from primary and secondary DCRs, the authorsfound no difference in the amount of inflammatory changesin the revision cases with or without previous intubation.Recurrence was explained by severe fibrotic changes at therhinostomy site, which was suggested to be related topostoperative tissue repair rather than the effect of siliconestent.34 Contrary to our anecdotal belief, the incidence or thetime to development of granulation tissue did not differsignificantly between patients with and without siliconeintubation in this study.

Success rates of nonintubated endoscopic DCR rangedfrom 74% to 92.3% at 8 to 49 months of follow-up.4e6,27,31

The success rate of nonintubated DCR in our study was

Table 7. Group Breakdown of Outcome in Patients with Bilateralor Unilateral Disease

Success Failure

Unilateral disease 106 (98.1%) 2 (1.9%)Bilateral disease 17 (85%) 3 (15%)

2144

95.3% at 12-month follow-up. This may be attributed toatraumatic creation of a large osteotomy2 with adequatesuperior bony clearance (at least 2 mm above internalpunctum), complete marsupialization of the lacrimal sac,20

maximal preservation of the nasal and sac mucosa,promoting edge-to-edge healing,11,12,25 and regular endo-scopic monitoring of ostial healing during the early post-operative period. Frequent endoscopic toileting withdebridement of “ostial-threatening” granulation tissue (Fig7AeF, available at http://aaojournal.org) may lead to animproved outcome that is yet to be verified in futureprospective trials.

In conclusion, patients with a history or subsequentdevelopment of contralateral PANDO (i.e., bilaterality) hada higher chance of failure (3/20) than those with unilateraldisease (2 of 108) (P¼0.005). Sodhi et al35 suggested youngage as a risk factor for failure of lacrimal drainage surgery.In our study, patients with failure tended to be younger,although this was not statistically significant (Table 4).Intraoperative endonasal access, bony clearance, size,mobility, and degree of marsupialization of the sac andimmunopathohistologic changes of nasal mucosa havebeen identified as possible predictors of outcome.20,36

Further prospective trials using these preoperative andintraoperative parameters will help to identify high-risksubgroups for operative adjuncts, for example, stent orantimetabolites in EEM-DCR.

References

1. Chandler PA. Dacryocystorhinostomy. Trans Am OphthalmolSoc 1936;34:240–63.

2. Goldberg RA. Endonasal dacryocystorhinostomy: is it reallyless successful? Arch Ophthalmol 2004;122:108–10.

3. Caversaccio M, Hausler R. Insertion of double bicanalicularsilicone tubes after endonasal dacryocystorhinostomy inlacrimal canalicular stenosis: a 10-year experience. ORLJ Otorhinolaryngol Relat Spec 2006;68:266–9.

4. Unlu HH, Toprak B, Aslan A, Guler C. Comparison of surgicaloutcomes in primary endoscopic dacryocystorhinostomy withand without silicone intubation. Ann Otol Rhinol Laryngol2002;111:704–9.

5. Mortimore S, Banhegy GY, Lancaster JL, Karkanevatos A.Endoscopic dacryocystorhinostomy without silicone stenting.J R Coll Surg Edinb 1999;44:371–3.

Chong et al � Randomized Trial of Intubation in Endoscopic DCR

6. Unlu HH, Ozturk F, Mutlu C, et al. Endoscopic dacryocys-torhinostomy without stents. Auris Nasus Larynx 2000;27:65–71.

7. Zuercher B, Tritten JJ, Friedrich JP, Monnier P. Analysis offunctional and anatomic success following endonasal dacryo-cystorhinostomy. Ann Otol Rhinol Laryngol 2011;120:231–8.

8. Madge SN, Selva D. Intubation in routine dacryocysto-rhinostomy: why we do what we do. Clin Experiment Oph-thalmol 2009;37:620–3.

9. Saiju R, Morse LJ, Weinberg D, et al. Prospective randomisedcomparison of external dacryocystorhinostomy with andwithout silicone intubation. Br J Ophthalmol 2009;93:1220–2.

10. Dulku S, Murray A, Durrani OM. Prospective randomisedcomparison of external dacryocystorhinostomy with andwithout silicone intubation: considerations of power. Br JOphthalmol 2011;95:151–2.

11. Tsirbas A, Wormald PJ. Endonasal dacryocystorhinostomywith mucosal flaps. Am J Ophthalmol 2003;135:76–83.

12. Tsirbas A, Wormald PJ. Mechanical endonasal dacryocysto-rhinostomy with mucosal flaps. Br J Ophthalmol 2003;87:43–7.

13. Woo KI, Maeng HS, Kim YD. Characteristics of intranasalstructures for endonasal dacryocystorhinostomy in Asians. AmJ Ophthalmol 2011;152:491–498 e1.

14. Javate R, Pamintuan F. Endoscopic radiofrequency-assisteddacryocystorhinostomy with double stent: a personal experi-ence. Orbit 2005;24:15–22.

15. Moore WM, Bentley CR, Olver JM. Functional and anatomicresults after two types of endoscopic endonasal dacryocysto-rhinostomy: surgical and holmium laser. Ophthalmology2002;109:1575–82.

16. Fayers T, Laverde T, Tay E, Olver JM. Lacrimal surgerysuccess after external dacryocystorhinostomy: functional andanatomical results using strict outcome criteria. Ophthal PlastReconstr Surg 2009;25:472–5.

17. Vagefi MR, Winn BJ, Lin CC, et al. Facial nerve injury duringexternal dacryocystorhinostomy. Ophthalmology 2009;116:585–90.

18. Woog JJ, Kennedy RH, Custer PL, et al. Endonasal dacryo-cystorhinostomy: a report by the American Academy ofOphthalmology. Ophthalmology 2001;108:2369–77.

19. Feretis M, Newton JR, Ram B, Green F. Comparison ofexternal and endonasal dacryocystorhinostomy. J LaryngolOtol 2009;123:315–9.

20. Davies MJ, Lee S, Lemke S, Ghabrial R. Predictors ofanatomical patency following primary endonasal dacryocys-torhinostomy: a pilot study. Orbit 2011;30:49–53.

21. Onerci M, Orhan M, Ogretmenoglu O, Irkec M. Long-termresults and reasons for failure of intranasal endoscopicdacryocystorhinostomy. Acta Otolaryngol 2000;120:319–22.

22. Mann BS, Wormald PJ. Endoscopic assessment of thedacryocystorhinostomy ostium after endoscopic surgery.Laryngoscope 2006;116:1172–4.

23. Ben Simon GJ, Brown C, McNab AA. Larger osteotomiesresult in larger ostia in external dacryocystorhinostomies. ArchFacial Plast Surg 2012;14:127–31.

24. Camara JG, Bengzon AU, Henson RD. The safety and efficacyof mitomycin C in endonasal endoscopic laser-assisteddacryocystorhinostomy. Ophthal Plast Reconstr Surg2000;16:114–8.

25. Codere F, Denton P, Corona J. Endonasal dacryocysto-rhinostomy: a modified technique with preservation of thenasal and lacrimal mucosa. Ophthal Plast Reconstr Surg2010;26:161–4.

26. Wu W, Cannon PS, Yan W, et al. Effects of Merogel coverageon wound healing and ostial patency in endonasal endoscopicdacryocystorhinostomy for primary chronic dacryocystitis.Eye (Lond) 2011;25:746–53.

27. Gibbs DC. New probe for the intubation of lacrimal canal-iculi with silicone rubber tubing. Br J Ophthalmol 1967;51:198.

28. Quickert MH, Dryden RM. Probes for intubation in lacrimaldrainage. Trans Am Acad Ophthalmol Otolaryngol 1970;74:431–3.

29. Anderson RL, Edwards JJ. Indications, complicationsand results with silicone stents. Ophthalmology 1979;86:1474–87.

30. Patrinely JR, Anderson RL. A review of lacrimal drainagesurgery. Ophthal Plast Reconstr Surg 1986;2:97–102.

31. Leong SC, Macewen CJ, White PS. A systematic review ofoutcomes after dacryocystorhinostomy in adults. Am J RhinolAllergy 2010;24:81–90.

32. Smirnov G, Tuomilehto H, Terasvirta M, et al. Silicone tubingis not necessary after primary endoscopic dacryocysto-rhinostomy: a prospective randomized study. Am J Rhinol2008;22:214–7.

33. Unlu HH, Gunhan K, Baser EF, Songu M. Long-termresults in endoscopic dacryocystorhinostomy: is intubationreally required? Otolaryngol Head Neck Surg 2009;140:589–95.

34. Ciftci F, Ersanli D, Civelek L, et al. Histopathologic changesin the lacrimal sac of dacryocystorhinostomy patients with andwithout silicone intubation. Ophthal Plast Reconstr Surg2005;21:59–64.

35. Sodhi PK, Verma L, Ratan SK. Young age - a risk factor forfailure of dacryocystorhinostomy. Orbit 2004;23:237–9.

36. Hammoudi DS, Tucker NA. Factors associated with outcomeof endonasal dacryocystorhinostomy. Ophthal Plast ReconstrSurg 2011;27:266–9.

Footnotes and Financial Disclosures

Originally received: September 12, 2012.Final revision: February 27, 2013.Accepted: February 27, 2013.Available online: May 11, 2013. Manuscript no. 2012-1401.1 Department of Ophthalmology and Visual Sciences, Prince of WalesHospital, Hong Kong, HKSAR, China.2 Department of Ophthalmology and Visual Science, The ChineseUniversity of Hong Kong, HKSAR, China.

Presented in part at: the American Academy of Ophthalmology meeting,October 16e19, 2010, Chicago, Illinois.

Financial Disclosure(s):The author(s) have no proprietary or commercial interest in any materialsdiscussed in this article.

Clinical trial registration number CUHK_CCT00159 (accessed April 1,2011); http://www.cct.cuhk.edu.hk/registry/publictrialrecord.aspx?trialid¼CUHK_CCT0015.

Correspondence:Kelvin K.L. Chong, MBChB(Hon), FCOphth, Department of Ophthal-mology and Visual Sciences, The Chinese University of Hong Kong, 4/FHong Kong Eye Hospital, Hong Kong. HKSAR, China. E-mail:chongkamlung@cuhk.edu.hk.

2145