Fig. 3.7 View into the wide open lacrimal sac on the right side 12 months after endonasal DCR (45¡ endoscope) with proof of function using the ßuorescein dye test. The uncinate process has been resected. A small remnant is seen just behind the upper portion of the lacrimal sac

External and endonasal (laser-assisted) lacrimal surgery have been compared in prospective and retrospective studies.

Mirza et al. retrospectively compared 49 patients with external DCR and 76 patients with endonasal laser-assisted DCR [54]. The success rates were 94% for external DCR with an average follow-up of 9 months and 64% for endonasal DCR after 12 months.

Hartikainen et al. compared external and endonasal laser assisted DCR in a prospective study [27]. Thirty two operations (all using silicone stenting) were reviewed 1 year post-operatively. Success was deÞned as objective demonstration of a patent lacrimal drainage system by irrigation and dacryoscintigraphy. Ninetyone percent of the lacrimal drainage systems operated on externally and 63% of those operated on endonasally were patent. The difference was statistically signiÞcant. The average operating time was considerably shorter with the endonasal procedure (23 vs. 78 min).

The problem with the comparision in these two papers is that the authors compared external DCR with laser assisted endonasal DCR which (at least at the moment) produces poorer results than conventional (cold) DCR.

More informative comparisons between the endonasal and external approaches were made by Dolman and Cokkeser et al. [16, 17].

In a retrospective analysis of 153 external and 201 endonasal DCR procedures, Dolman found comparable success rates of 90.2% and 89.1% (Òfull successÓ) respectively [17]. The average operating time was

signiÞcantly shorter for endonasal DCR (18.5 min vs. 34.3 min). Post-operative epistaxis requiring nasal packing occurred in 4.6% and 5.5% respectively. With external DCR local wound infection occurred in two cases and punctum eversion in six. With endonasal DCR there were periorbital lesions in Þve cases. Cokkeser et al. performed conventional external lacrimal surgery in 79 patients and an endonasal operation (hammer and chisel technique) in 36 patients (51 sides) [16]. In this prospective study, comparable success rates of 89.8% for the external procedure and 88.2% for the endonasal procedure were obtained after an average of 25 months. Success was deÞned as the absence of epiphora or other signs of chronic dacryocystitis and positive irrigation of the lacrimal system within the Þrst 6 months. The endonasal procedure was associated with fewer complications, lower morbidity and shorter average operating time (33 vs. 65 min).

Feretis et al. compared the outcome of external (90 patients) vs. endonasal (41 patients) DCR according to the ÒGlasgow beneÞt inventory questionnaireÓ and found no signiÞcant differences between the results [25].

In order to establish the value of endonasal DCR, Woog et al. performed a literature search of the years 1968Ð2000 on behalf of the American Academy of Ophthalmology and evaluated 64 papers [94]. On the basis of the analysis performed, it was difÞcult to arrive at evidence-based conclusions either on the efÞciency of the endonasal and external DCR procedures as a whole or on technical details (e.g. pre-operative workup, visualisation, surgical instruments, size and location of the neo-ostium, silicone splinting). In the end most questions remained open and further prospective studies were recommended. From the comparisons it was found, that the results of endonasal DCR were poorer than those of the external procedure. However, in most cases, the comparisons were with laser assisted procedures, which had poorer results. Nevertheless, the average success rates after endonasal DCR appear to be slightly lower than those after external DCR [52].

3.3Endonasal Endoscopic Laser Dacryocystorhinostomy (ELDCR)

Since its introduction in surgical practice, laser technology has improved the operative management of a number of procedures. In an attempt to achieve precise

Table 3.1 The advantages and disadvantages of the endonasal laser DCR over non-laser endonasal DCR

Advantages

Can be performed under local anaesthetic Can be performed on anticoagulated patients Shorter operative time

More effective haemostasis and low haemorrhage rates

Disadvantages

Lower success rate

Expensive equipment

Laser precautions required

bone removal with meticulous haemostasis, laser DCR was developed and Þrst described by Massaro et al. [51]. Since then, there have been a number of instances reported using various types of laser for DCR with variable results.

The type of laser appropriate for a DCR would allow delivery via ßexible optic Þbres, achieve effective bone ablation and provide good haemostasis with a relatively shallow depth of penetration. Therefore the potassium titanyl phosphate (KTP/532), diode and holmium:yttrium aluminium garnet (Ho:YAG) are suitable. The carbon dioxide (CO2) laser is not ideal because of its poor haemostatic properties, poor bone ablation and cumbersome delivery system. The Argon laser also has relatively poor bone ablation.

The Ho:YAG laser Þbres have multiple use speciÞcation and this can potentially reduce the cost per procedure. The major disadvantage is the splattering of tissue soiling the lens, requiring frequent cleaning and more collateral damage when compared to the KTP laser.

The KTP/532 with its star pulse mode is most suitable as it vaporises the bone effortlessly and without splattering. The diode laser also has sufÞcient power to ablate bone. The major disadvantage of the KTP and the Diode laser is that the optical Þbre is marketed for single use and therefore the cost per procedure for these lasers is signiÞcantly higher.

The endonasal laser DCR (ELDCR) is similar to the cold-instrument EDCR technique with the exception that laser energy is used to vaporise the mucosa and ablate the bone to create a Þstula. However, the success rates following non-laser EDCR are somewhat higher with a number of studies quoting success rates of over 90% [9, 85, 100]. The better surgical outcome with conventional surgery is related to a wider bony opening and avoiding the thermal damage caused by

the laser which produces more Þbrosis and occlusion at the rhinostomy site. The advantages and disadvantages of the laser technique are given in Table 3.1.

3.3.1 Indications for ELDCR

ELDCR is indicated for signiÞcant symptoms such as epiphora due to nasolacrimal duct obstruction (NLDO) that is not relieved by simple probing and syringing. It is not indicated for sole obstruction in the puncti, canaliculi, common duct and the lacrimal sac. In many patients there is some proximal obstruction associated with distal blockage. In such cases, gentle probing and dilatation in conjunction with a DCR and insertion of stents can be performed although the results of such an approach are not as favourable as cases of pure distal blockage. ELDCR can also be used in cases of acute dacryocystitis complicated by abscess formation [57]). ELDCR is ideal when a patient is anticoagulated or has a coagulopathy. Not only does it avoid any disruption to their anticoagulant therapy but it can also be done as a day case procedure. It is also more amenable to being performed under local anaesthetic than other techniques, which is useful in those patients unÞt for a general anaesthetic. Revision surgery after failed external or endonasal DCR or ELDCR can be performed effectively with the laser, removing the scarred tissue bloodlessly, although the outcome cannot be reliably predicted on account of excessive scarring which may involve the canaliculi and the common canaliculus. It is important that during the procedure the laser power setting is high enough to effect vaporisation rather than charring as the latter will mean that further laser energy and heat may be dissipated to surrounding tissues and produce more scar tissue.

Dacryoliths usually require removal via a wide opening of the lacrimal sac which can be achieved with the cold instrument EDCR (Fig. 3.1) or ELDCR.

3.3.2 Contraindications for ELDCR

The endonasal approach is inappropriate in the presence of malignant lesions of the lacrimal system or the surrounding tissues. In active WegenerÕs Granulomatosis,

any instrumentation induces marked adhesions and stenosis and a DCR by any method is contraindicated. A relative contraindication is a history of trauma as the bone medial to the sac may be thick.

3.3.3 Surgical Technique for ELDCR

One preferred technique is using the KTP/532 laser [55] on star pulse mode with a power setting of 50 W, 10 ms, 10 pulses/s for soft tissue and 70 W, 5 ms, 20 pulses/s for bone.

The patient is placed in the supine position at 15¡ reverse Trendelenburg. Appropriate laser safety precautions for the patient and operating team are taken to avoid ocular injury. Wet eyepads are placed over the patientÕs eyes. The operation can be performed under either local or general anaesthesia as per patient and surgeon preference.

The nasal operating site is accessed either with the microscope or the endoscope. When the operating microscope is used, the 300-mm objective is further away from the operating site and thus remains soilfree. It also provides useful magniÞcation. A KillianÕs speculum is placed in the nostril and the transilluminated site is located. The use of the microscope is however cumbersome and can add signiÞcantly to the operating time in inexperienced hands. Endoscopy with video monitoring is much more popular due to its superior visualisation and it is easier to manipulate, though the lens gets soiled due to smoke, blood and debris. Frequent cleaning is required, particularly when using the Holmium:YAG laser. The 0¡ endoscope is adequate in most cases although the 45¡ endoscope may afford a better view into the sac. Each

approach has its advantages and the choice would depend on individual training, preferences and the availability of equipment and dedicated instruments.

The upper punctum is dilated and the vitreo retinal light pipe is inserted. The pipe is advanced, initially in a vertical direction through the punctum for a millimetre or so, and then horizontally along the cannaliculus towards the medial canthus. Some resistance is then felt at the common cannaliculus (the soft stop) before it touches the mucosa of the medial wall of the sac (hard stop). From the hard stop, the pipe is withdrawn slightly and advanced in an inferior-medial slanting direction so that it passes into the lacrimal sac. The light is inserted into the upper punctum as it is easier to position the light pipe in the inferior part of the lacrimal sac. This helps to place the rhinostomy in a dependent position, avoiding the formation of lacrimal sump syndrome (a blind pouch with mucus collection of and recurrent dacryocystitis). Cannulating the upper canaliculus also prevents injury to the functionally more important inferior canaliculus. The light is then held in the most dependent position to show the position intranasally. It must be kept in the same position while the rhinostomy is being made to avoid Þring the laser in several positions when the light moves.

If the light pipe is accurately positioned in the lacrimal sac, it is usually seen as a bright and sharp spot underneath the tissues, just anterior to the attachment of the bony middle turbinate to the lateral nasal wall. The area of maximal brightness corresponds with the posterior end of the lacrimal sac where the overlying bone is thinnest, and not the centre of the sac. Another landmark for sac location is the maxillary line, a bony eminence which extends from the anterior attachment of the middle turbinate to the root of the inferior turbinate. It overlies the maxillary-lacrimal suture line

within the lacrimal fossa, the light of the endoscope may need to be reduced, to accurately visualise the spot [if there is an agger nasi cell]. The mucosa of the transilluminated area can be inÞltrated with 0.25 mL of 1% lignocaine with 1:200,000 adrenaline.

The laser optical Þbre is taken to the operation site through a handpiece. Some handpieces contain a second channel that is used to evacuate smoke and debris generated at the operation site. The distal end of the handpiece may be bent by about 25¡ so that the beam is directed laterally. For the KTP and Ho:YAG lasers, the laser probe is maintained in near contact mode during the procedure and the endoscope tip is positioned approximately 2Ð3 cm from the target site. The transilluminated area of mucosa covering the medial lacrimal bone is vaporised and the procedure is continued through the bone to make a shallow pit of about 4Ð5 mm in diameter. The transillumination becomes brighter as the bone is thinned. The process becomes difÞcult depending on the thickness of the bone formed by the nasofrontal process of the maxilla. Posterior to the anterior lacrimal crest, resection of the paper-thin lacrimal bone is easier. Bleeding can usually be arrested by using the laser in defocused mode. The vaporisation is continued until an opening of around 5Ð8 mm in diameter is created in the centre of the thinned-out bone. The laser cannot ablate charred tissue and so with continued use on such tissue, heat is dissipated through the surrounding tissues, increasing thermal injury.

The next step is to make an opening into the lacrimal sac. Movement of the light probe will conÞrm the location of the sac wall. The mucosa of the lacrimal sac is vapourised, again in the direction of transillumination. Alternatively, at this stage, a probe is passed into the lacrimal sac and the mucosa tented medially into the bony opening to conÞrm the location of the sac. On breaching the lacrimal sac mucosa a Þstula (rhinostomy) is created between the nasal cavity and the lacrimal sac. Various measuring devices can be used to assess the size of the rhinostomy to ensure uniformity. The rhinostomy should be at least 5 mm and preferably 10 mm in diameter to reduce the possibility of subsequent closure. Furthermore, the rhinostomy should be located as low as possible as a high rhinostomy results in a sump syndrome, predisposing to recurrent infections of the sac and the duct. The sac can be palpated through the rhinostomy opening to ensure that there is no pocketing of mucus or pus, or Þbrous strands, within the sac.

After removal of the light pipe, the lacrimal system is ßushed to ensure free ßow into the nasal cavity via the Þstula. Forcible syringing should be avoided. Unsuccessful syringing indicates an additional proximal obstruction that may have been overlooked at the initial assessment.

A silicone tubed stent is used by a number of surgeons to maintain the patency of the rhinostomy [54, 66, 78, 95]. One end of the tubing is inserted and fed through the upper canaliculus, and the other through the lower canaliculus, so that the ends come out of the new opening into the nasal cavity. The ends are held in situ by a Watzke sleeve or gently tied so that a small loop remains at the inner canthus. It is best to avoid excessive tightening of the loop as this may eventually cause Òcheese wiringÓ at the medial canthus. Stents are removed by cutting the loop at the inner canthus and pulling the sleeve with stents from the nostril.

3.3.4 Potential Problems with ELDCR

¥A narrow punctum requires repeated careful dilatation before the light pipe can be passed. SigniÞcant trauma to the punctum and the canaliculus will lead to scarring, obstruction and failure with persisting epiphora.

¥Intraoperative bleeding is rarely troublesome. Minor bleeding can be controlled with a few laser strikes in defocused mode. Any bleeding that is not easily controlled by a few laser strikes in a defocused mode should be controlled with the application of topical decongestants or vasoconstrictors on pledgets.

¥The reßection of the laser aiming beam from the tissues can be strong and this has the potential to be mistaken for the transilluminated light pipe beam with consequent vaporisation undertaken at the incorrect site. It is often helpful to intermittently point the aiming beam away from the operating site and check the position of the light pipe in the sac.

¥The transillumination spot is not always easy to locate. Its location may be aided by further manipulation of the light pipe. Even if the light is located, it may be diffuse rather than bright and sharply demarcated, due to a number of factors:

¥Hypertrophied anterior end of the middle turbinate