Prospective randomized comparison of external dacryocystorhinostomy and endonasal laser dacryocystorhinostomy

Prospective Randomized Comparison of External Dacryocystorhinostomy and Endonasal Laser Dacryocystorhinostomy Jouko Hurtikainen,

MD,’

Reich Grenman,

MD,

PhD,’ Pauli Puukka,

MSc,’ Heikki Seppgii, MD, PhD’

Objective and Design: The introduction of endonasal laser dactyocystorhinostomy (ENL-DCR) in the early 1990s showed great promise of changing dacryocystorhinostomy into an elegant, minimally invasive procedure from the traditional external dactyocystorhinostomy (EXT-DCR). This prospective, randomized study compares these two operations, their success rates, surgical durations, and postoperative symptoms. Participants: A total of 64 cases in 61 patients with primary acquired nasolacrimal sac or duct obstruction were divided into 2 subgroups by symptoms (simple epiphora and chronic dacryocystitis). These patients were randomized within both subgroups into 2 operation groups with 32 cases in each group. Altogether, 32 EXT-DCRs and 32 ENL-DCRs were performed. The silicone tube was removed Intervention: at 6 months after surgery. The final follow-up visit was at 1 year after surgery. The patency of the lacrimal passage was investigated by irrigation, and patients were questioned about their symptoms. Main Outcome Measures: The patency of the lacrimal passage to irrigation and the duration of surgery were measured. Results: The success rate at 1 year after surgery was 91% for EXT-DCR and 63% for ENL-DCR after primary surgery. The difference was statistically significant (P = 0.016). The surgical duration for ENL-DCR was three times shorter than for EXT-DCR, the average duration being 23 minutes and 78 minutes, respectively (P < 0.0001). Conclusions: The EXT-DCR, when compared with ENL-DCR, seems to provide superior operation results in primary acquired nasolacrimal duct obstruction. Ophthalmology 7998; 705:7 706- 7 7 73

The traditional treatment of nasolacrimal obstruction is an external dacryocystorhinostomy (EXT-DCR), often performed by an ophthalmologist. The technique of EXTDCR was described originally in 1904 by Toti’ and later modified by Dupuy-Dutemps and Bourguet* with the addition of suturing of the mucosal flaps, thus forming an epithelium-lined fistula. Success rates of 85% to 95% for EXT-DCR have been reported.“-* Disadvantages of EXTDCR include scarring of the facial skin, risk of copious hemorrhage, and disruption of medial canthal anatomy. The endonasal approach was introduced in 1893 by Caldwell’ but has been in limited use mainly because of

Originally Revision

received: accepted:

May 1, 1997. November 18, 1997.

’ Department

of Ophthalmology,

2 Department land.

of Otorhinolaryngology,

University

of Turku,

University

Turku,

of Turku,

Finland.

Turku,

Fin-

3 Social Insurance Institution, Research and Development Center, Turku, Finland. Presented in part at the IVth International Congress of the Internatlonal Society of Dacryology, Stockholm, Sweden, June 12, 1996. Supported in part by a grant Turku, Finland.

from

the Turku

University

Patients and Methods

Foundation,

The authors have no proprietary interest in any of the equipment mentioned in this article. Reprint requests to Jouko Hartikainen, MD, Department of Ophthalmology, Turku University Central Hospital, FIN-20520 Turku, Finland.

1106

the difficulty in visualization of the endonasal anatomy during surgery. Massaro, Gonnering, and Harris” in 1990 were the first to describe an endonasal dacryocystorhinostomy technique using laser (ENL-DCR) for creating the opening between the nasal cavity and the lacrimal sac. The intranasal operative site was visualized with an operating microscope, and the desired site of entry was identified by introducing a fiberoptic light needle into the lacrimal sac through the canaliculus. They used a highpower argon blue-green laser. Potassium titanyl phosphate (KTP),“,‘* carbon dioxide (CO,),” holmium:YAG (Ho:YAG),13-15 neodymium:YAG (Nd:YAG),” and combined C02-Nd:YAG16 lasers have been used with a video endoscope or operating microscope. Success rates of 68% to 85% have been reported.‘2-‘7 In this study, we report the results of a randomized prospective comparison between the traditional external DCR and endonasal laser DCR using combined C02Nd:YAG laser.

Patients referred to our clinic because of nasolacrimal obstruction were investigated using lacrimal irrigation and probing of the canaliculi with a blunt-tipped Bangerter lacrimal cannula up to bony contact of the lacrimal sac fossa. The lids were inspected, focusing on the positions of the lacrimal puncta and

Hartikainen

et al * Dacryocystorhinostomy Table

1. Patient

Data

External Characteristic

Total

No. of patients/operations

Dacryocystorhinostomy

EndonasalLaser Dacryocystorhinostomy

61164

32132

31132

65.0 ? 14.2 23-89

64.8 5 14.0 25-86

65.3 t 14.6 23-89

Age (yrs) Mean

? SD

Sex I’ Male/female Duration of symptoms (yrs) Mean Range Laterality of surgery (right/left) No. of patients with a history of acute dacryocysnns No. of patients with previous silicone intubation performed SD = standard

11/53

4128

9 l->50 28136

9 l->50 11/21

included

patients

were

8 l->50 17/15

7

2

5

1

1

0

deviation.

the function of the orbicularis muscle. Anterior rhinoscopy and Schirmer’s test were performed. Dacryocystography was performed on all patients at the department of radiology. Only patients with primary acquired nasolacrimal sac or duct obstruction and with a duration of symptoms for more than 1 year were included in this study. Exclusion criteria were canalicular or common canalicular obstruction ascertained with probing, noticeable lower lid laxity, previous lacrimal surgery with the exception of silicone intubation of nasolacrimal duct, age younger than 15 years, suspicion of malignancy, radiation therapy, posttraumatic bony deformity, and bone diseases. The

7125

divided

into

two

subgroups

by

symptoms: those who had simple epiphora with no discharge and those who had chronic dacryocystitis with purulent discharge. Altogether, 64 cases in 61 patients were randomized within both subgroups for either EXT-DCR or ENL-DCR. One patient with bilateral nasolacrimal obstruction was randomized for bilateral ENL-DCR, and two patients with bilateral nasolacrimal obstruction were randomized for both EXT-DCR and ENL-DCR. Thus, both groups included 32 cases. Of these cases, 11 were males and 53 were females. The mean age was 65 years (Tables 1,2). No preoperative selection based on the result of anterior rhinoscopy or dacryocystography was performed. All operations were performed between January and December 1994. The EXT-DCR and ENL-DCR operations were performed with the patient under general anesthesia. Cotton pledgets soaked with 200-mg cocaine (total dose) and epinephrine 1:10,000 in saline were applied intranasally between the posterior part of the middle and lower turbinate, as well as the roof of the nasal cavity, to achieve a good hemostasis. Intravenous antibiotics (cephuroxim, Zinacef, 750 mg; Glaxo Wellcome, Research Triangle Park, NJ) were given at the beginning of the anesthesia preceding surgical interventions. In both operations, a silicone tube (S 1.1000; Bika, F.C.I., Issy-Les-Moulineaux, France) was inserted and tied with several knots without fixating it to the nasal wall. Special care was taken not to tie the tube too tightly. On completion of the operations, antibiotic-corticosteroid (neomycin, polymyxin B and dexamethasone, Maxitrol; Alcon, Ft. Worth, TX) ointment was applied. After surgery, the patients used dexamethasone-chloramphenicol (Oftan DexaChlora, Santen, Tampere, Finland) drops three times daily for 2 weeks, and oral cephalexin (Kefexin, Orion, Espoo, Finland) 500 mg three times daily for 1 week. The EXT-DCR operations were performed by one ophthal-

mologist (JH) using the technique of Dupuy-Dutemps and Bourguet. A straight incision of approximately 2 cm in length was made medial to the angular vein, starting at the level of the medial canthal ligament. The orbicularis muscle fibers were separated with blunt dissection, and 4-O silk traction sutures were used to open the wound. The periosteum overlying and medial to the anterior lacrimal crest was exposed. The periosteum incision was made just medial and inferior to the bony insertion of the medial canthal tendon. The osteotomy, approximately 12 mm in diameter, was created with Kerrison rongeurs. The lacrimal sac and nasal mucosa were opened in a longitudinal fashion to form anterior and posterior flaps. The lacrimal sac mucosa was inspected and biopsied for pathologic study. Occasional lacrimal sac and nasolacrimal duct dacryoliths were removed. The site of nasolacrimal obstruction was localized by probing, and the internal pun&urn was inspected as well. The posterior flaps were sutured with two or three 5-O Dexon sutures (Davis-Geck, Wayne, NJ), and the silicone tube was inserted and tied with several knots. The anterior flaps were sutured with running 5-O Dexon. The periosteum and orbicularis muscle were closed in separate layers with 5-O Dexon sutures. The skin incision was closed with three 6-O Novafil (DavisGeck) sutures. Nasal packing was not performed. In the ENL-DCR operations, the lacrimal passages were probed and a 20-gauge fiberoptic light pipe of the type used in pars plana vitrectomies (Storz, Inc, St. Louis, MO) was lubricated with antibiotic ointment and inserted into the lacrimal sac through either of the lacrimal canaliculi into contact with the medial wall of the lacrimal sac. The light was visualized endonasally using an otorhinologic operating microscope (Zeiss OPMI- 1, Oberkochen, Germany), and 1% lidocaine-epinephrine (1 ml) was injected locally into the lateral wall of the nasal cavity, guided by the light pipe in the lacrimal sac. The eyes of the patient were protected from laser burn by covering them with several moistened cotton patch layers. The nasal mucosa was evaporated and a bony ostium was created, guided by the light pipe, using a continuous wave COZ-Nd:YAG combined laser (Lasermatic-Dual Laser, Helsinki, Finland), applying the laser beam endonasally in visual control using the microscope. A bony opening of approximately 10 X 5 to 7 mm was produced, beginning from the thin lacrimal bone at the inferomedial wall of the lacrimal sac fossa, where the light was visualized most easily, and then the opening was enlarged to the thick frontal process of the maxillary bone at the medial wall of the

1107

Ophthalmology Table

2. Data

Volume 105, Number of Patients

Divided

into

6, June 1998

Two

External Dacryocystorhinostomy Characteristic No. of patwnts/operations Age (yrs) Mean *SD Range Sex Male/female Duration of symptoms (yrs) Mean Range Laterahty of surgery (right/left) No. of panents wth a history of acute dacryocystltls SD = standard

Chrome Dacryocystitis

Simple Epiphora

16116

16116

Chronic

15116

16116

Dacryocystitis

a 64.6 k11.3 37-79

65.1 kl6.6 25-86

64.8 213.7 23-86

65.8 t15.9 34-89

2114

2/14

3113

4112

8 l-40 6110

10 l->50 5/11

0

8 l-40 1016 0

9 l->50 719 5

deviation.

sac. The power

settings of the laser varied

from 5 to 10 W for the CO1andwas5 W for Nd:YAG. Before the siliconeintubation, the lacrimal sacand the operatingfield were rinsedwith saline.No nasalpackingwas used.The ENLDCR operationswereperformedin cooperationby an otorhinolaryngologist (RG) and an ophthalmologist(JH). The ophthalmologistinsertedthe light pipe into the sacandassistedduring the laserablationwhile the otorhinolaryngologistperformedthe actual lasersurgery. The duration of surgery was measuredfrom the dilatation of lacrimal puncta and irrigation of the lacrimal systemto the end of the applicationof the pressurebandagefor the EXTDCR or to the applicationof antibiotic-corticosteroidointment for ENL-DCR. The first follow-up visit of the ENL-DCR group, 1 week after surgery,wasperformedby the otorhinolaryngologist.The intranasalrhinostomysitewascleanedfrom mucousanddebris. The first follow-up visit performedby the ophthalmologistin both groupswas 1 month after surgery.The operatedlacrimal passagewas irrigated with salinesolution,and the tightnessof the silicone tube was checked.At 6 monthsafter surgery,the siliconetube was removed and the lacrimal passagewas irrigated. The tube was removedby cutting the siliconetube betweenthe punctaandby eitherblowing the noseor by extracting the tube from the nosewith forcepsin anteriorrhinoscopy.The final follow-up visit was at 1 year after surgery. Patientswere asked,usinga questionnaire,about the presenceor absenceof dischargeand about watering of the eye outdoorsor indoors. The patientswho underwentEXT-DCR were questionedabout their satisfactionwith the cosmeticsof the cutaneousscar,and the site wasinspected.The patency of the lacrimalpassage was investigatedby irrigation and dacryoscintography. The statisticalanalyseswere performedusingSAS 6.04 statistical software (SAS Institute, Inc., Cary, NC). The group meanswere comparedwith Student’st test. Successratesand

1108

Endonasal Laser Dacryocystorhinostomy

Stmpk Eeiphora

lacrimal sacfossaup to the anteriorlacrimalcrest.The carbonized bone wasremovedwith the suctiontip. In mostcases,the wall of the lacrimal sac was openedseparatelywith the laser andbiopsiedfor pathologicstudy with endoscopicforceps.The assistingophthalmologistfacilitated the laser ablation of the lacrimal sac wall by tenting the medial sacwall nasally with either the light pipe or the Bangerterlacrimal cannulainserted into the lacrimal

Subgroups

other categoric variableswere testedwith Fisher’sexact test. Statistical significancesare given as exact probability values. The P valuesgreaterthan 0.05 are regardedas nonsignificant.

Results Thirty-two EXT-DCR operationsand 32 ENL-DCR operations were performedbetweenJanuaryandDeCember1994in a prospectively randomizedfashionon 61 unselectedpatientswith primary acquirednasolacrimalsacor duct obstruction.Complicationsduring surgeryseldomoccurred; there wasno copious intraoperative hemorrhage.After surgery, one patient in the EXT-DCR grouprequiredan anteriornasaltamponadeandhospitalization for 3 days. In the ENL-DCR group, there were no intraoperativeor postoperativehemorrhages.No decisionto convert the ENL-DCR operationto the EXT-DCR wasnecessary nor wasmiddleturbinectomyrequiredin either group.The averagedurationof surgerywas78 minutes(standarddeviation ? 12.8; range,60- 115minutes)for EXT-DCR and 23 minutes (standarddeviation 5 5.7; range, 14-38 minutes)for ENLDCR. The difference wasstatisticallysignificant(P < 0.0001). Ethmoid sinuseswere incorporatedwithin the osteotomyin 7 (22%) of 32 in the EXT-DCR group and in 21 (66%) of 32 in the ENL-DCR group. A tissuesamplefrom the lacrimal sac was obtainedin 32 (100%) of 32 in the EXT-DCR group and in 24 (75%) of 32 in the ENL-DCR group. Of the 24 samples in the ENL-DCR group, 6 were too small or affected by the laser,making histologicanalysisimpossible,whereasall samplesin the EXT-DCR group could be processedfor histologic analyses. The operationwasclassifiedas successfulby the objective demonstrationof a patent nasolacrimalsystemthrough irrigation. The successrate at 1 year after surgery, on the basisof the primary surgery,was29 (91%) of 32 for the EXT-DCR and 20 (63%) of 32 for the ENL-DCR group. The difference was statistically significant (P = 0.016). The patency rates at 1 month were 32 (100%) of 32 for the EXT-DCR and 31 (97%) of 32 for the ENL-DCR group. At 6 months,the patency rates were 32 (100%) of 32 and 25 (78%) of 32, respectively, as

Hurtikainen

et al * Dacryocystorhinostomy

Table 3. Patency Rates in Subgroups at Scheduled Postoperative Follow-up Vlslts External Dacryocystorhinostomy Postoperative Follow-up Visit 1 mo NO. W) 6 mos NO. (%I 12 mos (final visx) NO. (%) No. = number

SDllple Eprphora

Chronic Dacryocystitis

EndonasalLaser Dacryocystorhinostomy Simple Eprphora

Chronic Dacryocystitu

16/16 (100)

16116 (100)

16116 (100)

15/16 (94)

16116 (100)

16116 (100)

1406 (88)

11/16 (69)

14116 (88)

15116 (94)

lo/16

lo/16

(63)

(63)

of patients.

tested by irrigation immediately after removal of the silicone tube. Before randomization, the patients with primary acquired nasolacrimal sac or duct obstruction were divided into two subgroups by symptoms. In the EXT-DCR group, two of the three failures were in the group of simple epiphora and one was in the group of chronic dacryocystitis. In the ENL-DCR group, there were six failures in both subgroups (Table 3). When we compared the success rates between the two operations in subgroups, we did not find a statistically significant difference in the simple epiphora group (P = 0.220) nor in the chronic dacryocystitis group (P = 0.0829). Dacryoscintography was performed in 28 cases (88%) in the EXT-DCR group and in 17 cases (53%) in the ENL-DCR group. Of the cases that were classified as success by irrigation, dacryoscintigraphy was performed in 25 (86%) of 29 in the EXTDCR group and in 17 (85%) of 20 in the ENL-DCR group. Of the DCRs that were patent to irrigation and were studied with dacryoscintigraphy, 19 (76%) of 25 in the EXT-DCR group and 14 (82%) of 17 in the ENL-DCR group were found to be patent to dacryoscintigraphy. At the final postoperative follow-up visit, one patient (3%) in the EXT-DCR and seven patients (22%) in the ENL-DCR group reported having discharge, and this was stated on the questionnaire. All these patients were postoperative failures to irrigation. Watering indoors was reported by 2 patients (6%) in the EXT-DCR group and by 11 patients (34%) in the ENLDCR group. In the EXT-DCR group, 1 of the 2 patients, and in the ENL-DCR group, 1 of the 11 patients were found to have a patent rhinostomy to irrigation. Watering outdoors was reported by 5 cases (16%) in the EXT-DCR group and by 13 cases (41%) in the ENL-DCR group (Table 4). Four of the 5 patients in the EXT-DCR group and 1 of the 13 patients in the ENL-DCR group were patent to irrigation. Interestingly, two of the three patients in the EXT-DCR group who were classified as failures to irrigation were asymptomatic. Both of these patients had postoperative obstruction at the common canaliculus, ascertained by probing through the inferior canaliculus. However, the obstruction of one was only partial, obstructing the inferior canaliculus, whereas the superior canaliculus was patent. Dacryoscintigraphy, performed before the final follow-up visit, also showed patency in this patient. In the ENL-DCR group, all the patients who were failures to irrigation had symptoms. At the final postoperative follow-up visit, 31 of the 32 patients in the EXT-DCR group did not have any complaints about

the cosmetics of the cutaneous scar. One patient, a 25-year-old woman, reported color difference between the scar and the skin around the scar. No cheloid formation occurred. The removal of the silicone tube was scheduled at 6 months after surgery in both groups. In one patient (3%) in the EXTDCR group, the tube dislocated superiorly 3 months after surgery and was removed. In the ENL-DCR group, one patient (3%) had superior dislocation of the tube at 6 weeks after surgery. In this patient, the tube was repositioned by retracting it intranasally in anterior rhinoscopy. It was removed 6 months after surgery. In five patients, the tube was removed earlier than scheduled (range, 2-5 months) because of irritation or infection. The removal of the silicone tube was easy in 28 cases in the EXT-DCR group and in 26 cases in the ENL-DCR group. In four patients in the EXT-DCR group and in six patients in the ENL-DCR group, there were difficulties in removing the tube because of tightness at the rhinostomy site. In one patient in the EXT-DCR group and in four patients in the ENL-DCR group, the ophthalmologist was not able to remove the tube. These patients were sent to the otorhinolaryngologist, who was able to remove the tube from the EXT-DCR patient and from two of the ENL-DCR patients after endonasal enlargement of the rhinostomy. In two ENL-DCR patients, the knots were dislodged into the sac and the tube was removed at a secondary operation. Two of the four patients in the EXT-DCR group and five of the six patients in the ENL-DCR group with difficulties in removing the tube were postoperative failures. Laceration of the puncta was observed after surgery in four patients (13%) in the EXT-DCR group and in seven patients (22%) in the ENL-DCR group. All four patients in the EXTDCR group were successes, whereas only two of the seven patients in the ENL-DCR group were successes. In this study, three postoperative follow-up visits in the EXTDCR group and four follow-up visits in the ENL-DCR group were scheduled. However, additional visits before the scheduled visits became necessary in 5 patients (16%) in the EXT-DCR group and in 14 patients (44%) in the ENL-DCR group. The characteristics of patients with failure of primary surgery are summarized in Table 5.

Discussion The introduction

of ENL-DCR

in the early 1990s showed

great promise of changing dacryocystorhinostomy

into an

1109

Ophthalmology

Volume 105, Number 6, June 1998

Table 4. Symptoms at 12 Months Postoperatively by Questionnaire Irrigation

with Reference to Patency to

Endonasal Laser Dacryocystorhinostomy

External Dacryocystorhinostomy

Asymptomatic Watering indoors Watermg Discharge

outdoors

No. = number

NO.

f%)

Patent to Irngaml

27132 2132 5132 l/32

(84)

25127

(6) (16)

112 415 O/l

(3)

Table 5. Characterlstlcs

Age (yrs)

Primary Operation

Postoperative Problems

Watering Asymptomatrc

3/F 4/F 5/M

73 60 57

EXT-DCR EXT-DCR EXT-DCR ENL-DCR ENL-LICK

6/F

55

ENL-DCR

7/M

37 53

ENL-DCR ENL-DCR

l/F 2/F

E/F

F = female;

M = male; EXT-DCR by otorhinolaryngologlst.

t Removed

at endonasal

$ Removed

at external

1110

19132 11/32 13132

(59) (34) (41)

19/19 l/11 l/13 O/7

7132

(22)

of Primary Failures

Interval until Problems (mos)

= external

dacryocystorhmostomy;

ENL-DCR

endoscoplc dacryocystorhinostomy; ORL = otorhmolaryngologlst; No. = number: * Removed

legation

The EXT-DCR, with various modifications, consistently has yielded success rates of approximately 9O%,3-8 whereas the success rates of ENL-DCR using different wavelengths have been 68% to 85%.12-17 Tarbet and Custcr,6 in their recent retrospective study of 153 EXT-DCRs with various etiologies performed for more than 9 years and with a follow-up time from 2 weeks to 9 years, reported a success rate of 96% after primary operation confirmed by irrigation. Their success rate with primary EXT-DCR performed in patients with idiopathic nasolacrimal obstruction was 98%. The success rate with primary and revised EXT-DCRs with various etiologies was 95%, showing a slight decline with increased duration of follow-up, the success rate being 89% at 1 year after surgery. The ENL-DCR has been performed using different

AsymptomatIc Watermg, discharge Watermg, lrrltatlon Copious discharge, irritation, pam Discharge, irritation, punctal laceration Discharge Watermg, tube drawn mslde the sac Watermg Discharge, lrrltanon Watermg Watering, discharge Discharge, nwtatlon Watering

ENL-DCR ENL-DCR ENL-DCR ENL-DCR ENL-DCR ENL-DCR ENL-DCR

9/F 10/F 11/M 12/F 13/F 14/F 15/M

f%)

of patients.

elegant, minimally invasive procedure from “the operation of bone cracking, cursing and postoperative failure.” ‘* The advantages of ENL-DCR mentioned in the literature include limitation of tissue injury to the discrete rhinostomy site, avoidance of a skin incision, excellent hemostasis, the ability to perform a lacrimal bypass operation on an outpatient basis with quicker rehabilitation and decreased overall healthcare expense, and patient preference.” The success rates for ENL-DCR12-” seem to have been lower than those reported previously for the traditional EXT-DCR,3-8 but no randomized studies to compart these operations have, to our knowledge, so far been published. In the current study, we have made a prospective, randomized comparison of the traditional EXT-DCR and ENL-DCR.

Failure NoJGender

Patent to NO.

endoscoplc

dacryocystorhmostomy.

dacryocystorhinostomy.

Time of Tube Removal bos)

Difficulties with Tube Removal (Yes/No)

Laceration of Punctum (Yes/No)

Type of Obstruction

Postsaccal

Yes Yes No Yes/ORL* No

No No No No

Presaccal Presaccal Postsaccal Postsaccal

3

No

No

Postsaccal

4 5

No NO

Yes Yes

Postsaccal Postsaccal

Yes/EESC-DCRt Yes No NO Yes Yea/EXT-DCRt: No

No No Yes No No NO Yes

Postsaccal Postsaccal Presaccal Postsaccal Presaccal Postsaccal Postsaccal

= en d onasal laser dacryocystorhinostomy;

Yes

EESC-DCR

= cndonasal

Hartikainen

et al * Dacryocystorhinostomy

wavelength lasers, beginning with the argon blue-green laser” with wavelengths of 488 and 514 nm, and later the CO2 laser” and the KTP laser11S’2with wavelengths of 10,600 and 532.0 nm, respectively. The Ho:YAG laser13”4 with a wavelength of 2100 nm, and later, in combination with the contact Nd:YAG laser15 with a wavelength of 1064 nm, has been used. In our pilot study of ENL-DCR, we used a combined CO,-Nd:YAG laser.16 The success rates have varied from 68% to 85% with follow-up times ranging from 6 weeks to 23 months.‘*-” Metson and associates14 have achieved the highest success rate, 85%, with a follow-up time from 7 to 23 months (mean, 13.2 months) using the Ho:YAG laser. Histologic findings in early primary acquired nasolacrimal sac or duct obstruction suggest that the obstruction may be a reversible process.lg Therefore, the inclusion criterion was the duration of lacrimal symptoms for at least 1 year. The patients with symptoms for less than 1 year were treated primarily with silicone intubation. In our study, the success rate was defined by an anatomically patent nasolacrimal system ascertained by irrigation at 1 year after surgery in all patients. Our success rates showed a similar tendency to previously reported success rates.3-8”2-‘7 The EXT-DCR had a primary success rate of 9 1% evaluated at 1 year after surgery, whereas the primary success rate of ENL-DCR was 63%. The success rate of ENL-DCR was slightly lower than reported previously.“-” It compares well with the success rate in our pilot study of endonasal CO,-Nd:YAG laser DCR, in which the success rate of 12 patients at 1 year after surgery was 67% after a single attempt and 83% after reoperation.‘” In our pilot study, we had problems with the breaking and extrusion of the tube in 4 of the 12 cases. This problem was alleviated by the use of dulledged forceps when tying the intranasal knot. The other problem in our pilot study was the laceration of the puncturn in 3 of the 12 cases caused by excessive tightening of the tube after spontaneous nasal displacement of the piece of silicon catheter surrounding the tube at the site of rhinostomy. In the current study, we did not use a double stent. Rather, we plaited the silicone tube with several knots to create a thicker stent at the rhinostomy site. This led to a new problem, because some of the knots dislodged into the lacrimal sac, making the removal of the tube by pulling either difficult or impossible. Forced pulling and endonasal surgical enlargement of the rhinostomy site were mechanical traumas that later led to the development of a scar membrane and postoperative failure in one of the four patients with difficulties in tube removal in the EXT-DCR group and in four of the six patients in the ENL-DCR group. Technically, these two operations represent quite different types of surgery. The EXT-DCR is a traditional operation with knife, scissors, forceps, and rongeurs, whereas ENL-DCR is a recent operation, the most significant instrument being the laser beam. We performed EXT-DCR with the technique of Dupuy-Dutemps and Bourguet, in which the anterior and the posterior flaps are sutured to achieve a controlled epithelium-lined anastomosis. The advantage of EXT-DCR was good intraoper-

ative visibility inside the lacrimal sac, allowing inspection of the internal punctum and lacrimal sac mucosa and permitting biopsies to be performed easily. Dacryoliths could be detected and removed also. The difficulties in this operation included intraoperative diffuse bleeding obscuring visibility, usually at the beginning of the operation, and a narrow space in which to suture the posterior flaps, although we created quite a large bony ostium. Advantages of ENL-DCR included the avoidance of a cutaneous wound, excellent hemostasis, both intraoperatively and postoperatively, and easy access to the rhinostomy site leading to short operative time. Disadvantages were the inability to detect possible lacrimal sac pathology, especially the site of the internal punctum, when performing the surgery without an endoscope, and the difficulty of performing a reliable biopsy in nearly half of the operations. Nuss and associates” have investigated the optimum wavelength for bony ablation. They noted that the continuous-wave CO2 laser provided effective bone removal, but both the continuous wave Nd:YAG and CO* lasers were much less precise than pulsed lasers, causing more peripheral thermal injury to nontarget tissues. In clinical use, we have found that the combined CO,-Nd:YAG laser provides an effective vaporization of tissue, including bone, and excellent hemostasis. We tried to avoid possible thermal damage to the lateral wall of the lacrimal sac by tenting the lasered medial sac wall nasally with either the light pipe or Bangerter lacrimal cannula inserted into the lacrimal sac. The postoperative harmful influence of peripheral thermal damage around the lasered rhinostomy site was minimized by careful removal of carbonized tissue with the suction tip and rinsing of the operating field with saline at the end of the operation. Woog and associates’3 pointed out four technical modifications resulting in an improved success rate for ENL-DCR: (1) creation of a larger DCR ostium (diameter, 6-8 mm), (2) removal of larger amounts of mucosa from the medial wall of the lacrimal sac and the superior portion of the nasal lacrimal duct in an attempt to visualize the common internal punctum, (3) use of a small drill for removal of char and bone in addition to laser ablation of nasal and lacrimal mucosa, and (4) use of a double stent. We agree with modifications (1) and (2), but the other modifications require comment. The combined C02-Nd:YAG laser provides an effective bone ablation, even in the area of the thick frontal process of maxillary bone at the nasal wall of the lacrimal sac fossa that, combined with the mechanical removal of carbonized tissue with the suction tip and tenting the lacrimal sac wall nasally during laser ablation, makes the use of a drill unnecessary. The use of a double stent in our pilot study led to laceration of the puncta from inadvertent tightening of the tube, resulting from nasal displacement of the outer stent. Therefore, we did not use it in the current study. After the secondary operations, as well as after additional ENLDCRs, we reached the conclusion that the most important modification resulting in an improved success rate for ENL-DCR is frequent postoperative intranasal cleaning of debris and mucus at the rhinostomy site performed by 1111

Ophthalmology

Volume

an otorhinolaryngologist. The postoperative care of the patient who underwent EXT-DCR is simpler, generally requiring either one or two follow-up visits, the last visit including silicone tube removal. There is a difference of opinion as to whether a silicone tube should be inserted in the EXT-DCR operation. A silicone tube, although inert material, may cause peripunctal granulation and chronic infection, canalicular laceration, dislocate superiorly, and may be cut by the uninitiated.” Allen and Berlin22 reported a statistically significant increase in the failure rate of primary EXT-DCR with silicone tubes inserted. However, Walland and Rose2’ did not find a significant difference in the rate of failure or postoperative infection in EXT-DCR cases with or without silicone intubation. Comparisons between silicone tubes from different manufacturers have not been performed. The differences may be caused by granuloma inducing impurities in some of the products. Silicone intubation has been recommended in cases involving common canalicular scarring, a large valve of Rosenmueller requiring surgical excision, or a small contracted or scarred lacrimal sac.22 Silicone tubing may be considered necessary in EXT-DCR cases in which none or only one of the flaps is sutured or to compensate for insufficient surgical skills. Our opinion is that the best epithelialized anastomosis is achieved by suturing the posterior and anterior flaps. The necessity of silicone intubation in ENL-DCR operations has not been questioned. The retention time of the silicone tube has varied from 6 days to 7 months.‘0-‘7 Boush and associates” have pointed out the importance of adequate silicone tube retention. In their study of 46 ENL-DCR cases, the success rate with silicone tube retention of 4 months, as planned, was 81%, whereas the success rate with premature silicone extubation was 22%. In the current study, we performed silicone intubation in both the EXT-DCR group and the ENL-DCR group to make the groups similar in this respect. Silicone extubation was planned to be performed 6 months after surgery. In five cases in the ENL-DCR group, the tube was removed prematurely because of infection or irritation, probably because of tightness of the tube. Although these cases were failures to irrigation after primary operation, the reason for the failure probably was not the premature extubation but rather the infection or irritation itself. Before surgery, we divided all the patients referred to our clinic because of nasolacrimal sac or duct obstruction into two subgroups by symptoms: those who had simple epiphora with no discharge and those who had chronic dacryocystitis with purulent discharge. Walland and Rose23 reported a soft tissue infection occurring in approximately 8% of patients after open lacrimal surgery. They found that a five-fold reduction in this rate can be achieved with routine administration of antibiotics after surgery. In the current study, all patients received prophylactic intravenous cephuroxim peroperatively and cephalexin 500 mg three times daily orally for a week after surgery. This antibiotic regimen satisfactorily covers the bacteriology of lacrimal duct obstruction in adults, which mostly consists of gram-positive organisms, Staphylococ-

1112

105, Number 6, June 1998 cus epidermidis being the most frequently isolated species. In patients with chronic dacryocystitis, gram-negative organisms were isolated in 26% of samples from the lacrimal sac, whereas they were isolated in only one sample (3%) of cases of nasolacrimal obstruction with simple epiphora.24 In the current study, there were no wound infections in the EXT-DCR group. The success rates did not appear to correlate with the presence or absence of discharge. The duration of EXT-DCR was 78 minutes on average, which is twice the duration Bartley25 reports in his article. Tarbet and Custer6 reported a significant decline in surgical time occurring with increasing physician-surgical team experience: the duration of simple EXT-DCR was, on average, 100 minutes in 1988, declining to 52 minutes in 1992. We have experienced the same tendency with further EXT-DCRs. The duration of ENL-DCR was three times shorter than the duration of EXT-DCR, being 23 minutes on average. However, it was even four to five times shorter than the durations of ENL-DCRs reported recently by others.‘2*‘4 The average surgical durations reported in previous articles using Ho:YAG and KTP lasers have varied from 89 to 116 minutes.‘2.‘4 In five articles on studies using Ho:YAG, contact Nd:YAG, CO*, KTP, and argon lasers, no operative time has been reported. 10.11.13,1s,17 In the current study, we have compared EXT-DCR and ENL-DCR in cases of primary acquired nasolacrimal sac or duct obstruction, The most important result in the comparison of the two different lacrimal operations is the success rate; the duration of surgery becomes important only when the two operations achieve comparable success rates. Currently, EXT-DCR is a superior operation, needing only few postoperative follow-up visits and with a very small risk of a disturbing cutaneous scar. Although ENL-DCR achieved a significantly lower success rate, it is preferred by many patients mostly because of the lack of a cutaneous wound. With developing techniques and experience, the success rate of ENL-DCR will increase, but more frequent postoperative follow-up visits for cleaning mucus and debris intranasally at the site of the rhinostomy are needed.

References 1. Toti A. Nuovo metodo conservatore di cura radicale delle suppurazioni croniche de1 sacco lacrimale (Dacriocistorinostomia). Clin Mod Firenze 1904; 10:385-7. 2. Dupuy-Dutemps, Bourguet. Pro&de plastique de dacryocysto-rhinostomie et ses resultats. Ann Ocul 1921; 158: 241-61. 3. Rosen N, Mordechai S, Moverman DC, Rosner M. Dacryocystorhinostomy with silicone tubes: evaluation of 253 cases. Ophthalmic Surg 1989;20:115-9. 4. Vanhoucke K, Colla B, Missotten L. Dacryocystorhinostomy: indications, operations, results and some variants. Bull Sot Belge Ophthalmol 1990;238:103- 10. 5. Dresner SC, Klussman KG, Meyer DR, Linberg JV. Outpatient dacryocystorhinostomy. Ophthalmic Surg 1991; 22: 222-4.

Hartikainen et al * Dacryocystorhinostomy 6. Tarbet KJ, Custer PL. External dacryocystorhinostomy: surgical success, patient satisfaction, and economic cost. Ophthalmology 1995; 102:1065-70. 7. Becker BB. Dacryocystorhinostomy without flaps. Ophthalmic Surg 1988; 19:419-27. 8. Emmerich KH, Busse H, Meyer-Rusenberg HW. Dacryocystorhinostomia externa. Ophthalmologe 1994; 9 1:395 -8. 9. Caldwell GW. Two new operations for obstruction of the nasal duct. N Y Med J 1893;57:581-2. 10. Massaro BM, Gonnering RS, Harris GJ. Endonasal laser dacryocystorhinostomy: a new approach to nasolacrimal duct obstruction. Arch Ophthalmol 1990; 108:1172-6. 1I. Gonnering RS, Lyon DB, Fisher JC. Endoscopic laser-assisted lacrimal surgery. Am J Ophthalmol 1991; 1I 1:1527. 12. Reifler DM. Results of endoscopic KTP laser-assisted dacryocystorhinostomy. Ophthal Plast Reconstr Surg 1993; 9:23 l-6. 13. Woog JJ, Metson R, Puliafito CA. Holmium: YAG endonasal laser dacryocystorhinostomy. Am J Ophthalmol 1993;116:1-10. 14. Metson R, Woog JJ, Puliafito CA. Endoscopic laser dacryocystorhinostomy. Laryngoscope 1994; 104:269-74. 15. Kong YT, Kim TI, Kong BW. A report of 131 cases of endoscopic laser lacrimal surgery. Ophthalmology 1994; 101:1793-800. 16. Seppti H, Grenman R, Hartikainen J. Endonasal C02-

17. 18.

19. 20. 21. 22. 23. 24. 25.

Nd:YAG laser dacryocystorhinostomy. Acta Ophthalmol (Copenh) 1994;72:703-6. Boush GA, Lemke BN, Dortzbach RK. Results of endonasal laser-assisted dacryocystorhinostomy. Ophthalmology 1994; 101:955-g. Dryden RM, Wulc AE. Surgery of the lacrimal system. In: Waltman SR, Keates RH, Hoyt GS, Frueh BR, Herschler J, Carroll DM, eds. Surgery of the Eye. New York: Churchill Livingstone, 1988; v. 2, chap. 50. Linberg JV, McCormick SA. Primary acquired nasolacrima1 duct obstruction: a clinicopathologic report and biopsy technique. Ophthalmology 1986;93: 1055-63. Nuss RC, Fabian RL, Sakar R, Puliafito CA. Infrared laser bone ablation. Lasers Surg Med 1988;8:381-91. Walland MJ, Rose GE. The effect of silicone intubation on failure and infection rates after dacryocystorhinostomy. Ophthalmic Surg 1994;25:597-600. Allen K, Berlin AJ. Dacryocystorhinostomy failure: association with nasolacrimal silicone intubation. Ophthalmic Surg 1989;20:486-9. Walland MJ, Rose GE. Soft tissue infections after open lacrimal surgery. Ophthalmology 1994; 10 1:608- 11. Hartikainen J, Lehtonen O-P, Saari KM. Bacteriology of lacrimal duct obstruction in adults. Br J Ophthalmol 1997;81:37-40. Bartley GB. The pros and cons of laser dacryocystorhinostomy. Am J Ophthalmol 1994; 117:103-6.