Simulator training (Weiss, OPRS 2009) or apprenticeship with an experienced surgeon might be appropriate for surgeons beginning to use these techniques.
- EndoDCR is most easily performed under general anesthesia. An oral endotracheal tube (ET tube) interferes least with the surgeon's hands and instruments, but a standard ET tube or a laryngeal mask airway (LMA) is also usually satisfactory. Pooling of blood in the posterior pharynx is likely and should be discussed with anesthesiologist.
- Several papers describe successful local anesthesia alone for endoDCR (see below). However, the lack of airway protection raises concern in the event of significant intraop bleeding.
- Local infiltration using lidocaine with epinephrine of the region of the uncinate process of the lateral wall of the nasal cavity from the anterior root of the middle turbinate inferiorly for about 2 cm under direct headlight observation or endoscopy (1.5‑inch 25‑gauge needle or 22‑gauge spinal needle work well). This is followed by placement of nasal packing soaked in 0.5% oxymetazoline (two 0.5" x 3" cottonoids usually suffice per side) for 5–7 minutes.
No prep is necessary, but standard draping is helpful. Commonly used instruments include 30‑degree endoscope/camera/video tower, Frasier suction, nasal speculum (medium), Freer or Cottle elevator, Kerrison rongeurs (small and medium), small Takahashi or Blakesley forceps, 4‑mm osteotome and mallet, 66 Beaver blade or sickle knife or 12 blade on long handle, punctal dilator, Bowman probes, silicone intubation of surgeon's choice.
For right-handed surgeons, both sides are most easily accessible by standing at the patient's right shoulder. This position is easier because the nondominant hand holds the endoscope or nasal speculum, leaving the dominant hand free for surgical instruments. The Mayo stand can be positioned at the head of the bed or over the patient, but in this location is at risk for contamination by the surgeon leaning over to look up the nose with the headlight. The video screen is positioned over the patient's left shoulder. Suction and endoscope cables can be at the patient's chest or on the Mayo stand at the head of the bed. Reverse for left-handed surgeons.
Depending on variability in anatomy and intraoperative bleeding, the surgery can be easier either with the view using an endoscope or the view directly with a headlight. Surgeons can readily have both options available to them for their cases.
Variations in the space around the middle turbinate and uncinate process may affect surgical technique. See Figures 1 and 2 for examples of narrow and wide middle meatus.
- With headlight observation and nasal speculum or the endoscope, incise nasal mucosa with Cottle or Freer about 0.5 cm anterior to the uncinate process and the root of the middle turbinate, starting superiorly and extending inferiorly along the path of the lacrimal sac for about 1.5 cm. Try to avoid injury to the nasal septal mucosa from the speculum or other instruments. The endoscope can be used for primary observation or at any time to double check anatomy.
- Still using headlight observation and nasal speculum or the endoscope, Kerrison rongeurs are positioned to bite the bone and mucosa just anterior to the middle turbinate, thereby exposing the lacrimal sac. Typically the deep end will pass lateral to or under the middle turbinate (which might require infracture) and the proximal cutting end will be positioned along the nasal mucosal incision. Multiple bites are taken, removing bone and mucosa and almost always exposing the lacrimal sac, which is then further exposed for all or almost all of its length. Bleeding during this rhinostomy can occasionally be challenging; repacking for a few minutes with cottonoids is usually adequate. For assistance with the anatomy, a retinal light pipe positioned in the lacrimal sac might be helpful for the first few cases or as desired for difficult cases. See article by Shams for anatomy of the lateral nasal wall (Shams, Curr Opin Ophthalmol 2015).
- Although infracture of the middle turbinate is occasionally necessary, it is generally wise to avoid injury to the turbinates when possible (Gustafson, OPRS 1999) because adhesions, bleeding, and damage to the ostiomeatal complex with subsequent maxillary sinusitis can occur.
- Gentle palpation in the region of the medial canthus at the lacrimal sac will now be seen from inside the nose as bulging in the region of the osteotomy. Sometimes this is easily seen by headlight observation, but endoscopic observation is often needed. Continued bone removal to better expose the sac is then performed.
- Depending on the remaining bone, useful instruments might include the Takahashi forceps, the Cottle elevator, and the Kerrisons. Switching back and forth between direct headlight observation and endoscopic observation might be helpful. Endpoint is exposure of at least 1 cm vertically of the sac, with the superior extent of the osteotomy above the valve of Rosenmuller (checked by passing a Bowman probe horizontally into the sac from the inferior punctum). The superior-anterior edge of the osteotomy can be the most difficult due to the thickness of the bone, the awkward handling of available tools, and the concern over injury to the cribriform plate and CSF leak. The Malhotra bone nibbler (Altomed) can be helpful (Patel, OPRS 2011). In rare cases, it might be necessary to remove this bone with cautious use of the 4‑mm osteotome; care must be taken to avoid superior extension of bone fracture to the cribriform plate.
- Incise the medial wall of the lacrimal sac with the 66 Beaver or sickle knife or 12 blade for the entire vertical length of the osteotomy, positioning the incision closer to the posterior edge of the osteotomy. This location permits easier removal of a central portion of the medial sac wall for biopsy or to reduce potential for the sac to reform. Remember that the blade must penetrate the periosteum around the sac as well as the mucosa and some force is needed. A Bowman probe can be used to confirm completion of the mucosal incision. Release of purulence is also reassuring.
- Examine the inside of the sac with the endoscope (30‑degree might be more useful here than 0‑degree) to look for typical rugae, mucosal integrity, and the occasional misplaced intracanalicular plug. If present, dacryoliths can be removed by pushing them medially with a Bowman probe or by Frazier suction or by using the Cottle or Takahashi.
- The anterior and posterior sac flaps can be subtotally removed using the small Kerrison rongeurs without any twisting motion and biopsied as needed. Removal of the flaps does not negatively impact success (Hodgson, OPRS 2014) and also serves to make reformation of the sac or sump syndrome potentially less likely.
- If desired, mitomycin C can be applied with standard techniques (see below).
- If desired, silicone intubation is performed using the surgeon's preference of stents. However, avoid using the double diameter Stent-tube because canalicular trauma is common. A Sisler lacrimal trephine (Visitec) is useful for stenosis at the valve of Rosenmuller if needed. Avoiding any tension at the puncta, the ends of the silicone stent can be tied or first passed through a silicone button and then tied.
- Hemostasis is most often adequate from the epinephrine and oxymetazoline, so that cautery and procoagulants are not needed. Even brisk bleeding often responds to reapplication of the oxymetazoline cottonoids with pressure for a few minutes.
- If there is significant iatrogenic trauma to the middle turbinate or nasal septum, consider placement of an absorbable fibrillar nasal tampon to prevent intranasal adhesions. Alternatively, use other nasal packing of the surgeon's choice that will be removed (see below in complications).
- Laser-assisted endoDCR: Reports range from 65% success (Tripathi, Eye 2002) to 96% success (Velegrakis, Am J Otolaryngol 2002).
- Moore compared endoscopic DCR to Holmium:YAG laser-assisted endoscopic DCR and found success rates of 71% with the laser and 86% without it (Moore, Ophthalmology 2002) whereas Maini showed no difference between the results from the two techniques (Maini, J Laryngol Otol 2007).
- Lasers studied include holmium, potassium titanyl phosphate, carbon dioxide, and others.
- Advantages could include improved hemostasis, but lasers might be less successful due to the often smaller ostium created as well as due to thermal-induced tissue damage and fibrosis leading to subacute failure.
- Mitomycin C application to the ostium at surgery: Routine mitomycin C use presents challenges related to its safe disposal and mutagenicity.
- Mitomycin C is a natural product isolated from Streptomyces species that reduces fibrosis and mucosal proliferation after topical application. Early reports used it for 30 minutes at 0.2 mg/ml (Kao, Ophthalmology 1997) and 0.5 mg/ml for 2.5 minutes (Ugurbas, Ophthalmic Surg Lasers 1997) although many other protocols have since been reported (Marcet, Curr Opin Ophthalmol 2014).
- Laser-assisted endoDCR plus mitomycin C: 99% success among 123 eyes after at least 30 months (Camara, OPRS 2000)
- Nonlaser endoDCR plus mitomycin C: 96% success among 110 eyes after at least 6 months (Kamal, OPRS 2014); similarly, 95% success in 224 eyes at 18 months (Dolmetsch, Ophthalmology 2010).
- Although these results are impressive, a study suggests that mitomycin C might be more useful in external DCR than endoDCR (Xue, Orbit 2014).
- Balloon-assisted endoDCR: Balloons and inflators add expense and the indications are unclear.
- 9-mm endonasal balloon: 92% success in 97 cases (Figure 3) (Silbert, Orbit 2010)
- 5-mm endocanalicular balloon: 91% success in 35 cases under local anesthesia (Ragab, Otolaryngol Head Neck Surg 2011); risk of iatrogenic canalicular damage and subsequent obstruction has limited adoption of this technology (Figure 4).
Figure 3. 9-mm endonasal balloon positioned in DCR ostium.
Figure 4. 5-mm endocanalicular balloon positioned at DCR ostium.
- Powered endoDCR
- Guarded electric burrs for osteotomy: 93% success in 283 cases (Ali, OPRS 2015)
- Can be useful in removing the thicker bone anteriorly
- Stentless endoscopic DCR. Routine use of stents is very common, but there is evidence that they might not be necessary. A number of conflicting studies have been published.
- In a randomized series of 128 eyes, success rates were higher in the stentless group of 65 eyes (89%) versus the stent group of 63 (57%) at 33 months (Mohamed, Ann R Coll Surg Engl 2013).
- Another study demonstrated no difference among 120 randomized patients (Chong, Ophthalmology 2013) and a third demonstrated 91% success among 132 cases without stents at 12 months and no new consecutive canalicular obstruction (Cannon, Ophthalmology 2013).
- A prospective, randomized, controlled interventional trial of 300 patients showed a statistically higher success rate in endonasal DCR surgery with tubes compared with no tubes. Failure rate was more than as twice as high when tubes were not used (12.2% vs. 5.3%) (Fayers, Ophthalmology 2016).