Statistics from Altmetric.com
Editor,—Canalisation of the nasolacrimal apparatus usually occurs at the same time throughout its length. However, its distal end has been shown to be occluded by a membrane in 73% of otherwise normal stillborn fetuses at term.
For typical dacryocystocoeles, a regimen of warm compresses and massage, with regular ophthalmological review to check for the first signs of dacryocystitis, seems to be reasonable. Should dacryocystitis supervene, the child should be admitted to hospital for the intravenous administration of antibiotics and probing of the nasolacrimal apparatus. Should the dacryocystocoele recur or epiphora ensue, and repeated probing does not give the result, it may be necessary to intubate the nasolacrimal apparatus or perform a dacryocystorhinostomy. In 1893, Caldwell described the first case of an endonasal operative approach to the lacrimal system. This technique was later modified by West and supported by Mosher in 1921. In spite of these attempts, the external dacryocystorhinostomy (DCR)—the technique inaugurated by Toti in 1904—was, for a long time, the most accepted procedure for lacrimal sac surgery. The reason for this was presumably limited transnasal visualisation caused by bleeding during endonasal dacryocystorhinostomy. In 1974, Jokinen and Karja revived the endonasal approach. Heerman and Neues used a microscope for a transnasal approach to the lacrimal sac, whereas McDonough and Meiring were the first to advocate endonasal endoscopic dacryocystorhinostomy (EEDCR), in 1989.
Using new instrumentation and techniques for endoscopic sinus surgery in general, many authors have proved that EEDCR can be performed with lower morbidity in adults and with success rates equal to those achieved with the traditional external approach.
A 4 month old girl presented with huge recurrent abscesses in the left medial canthal region (Fig 1) and a huge epiphora. Initially she was managed by conservative methods (warm compresses and massage over the swollen lacrimal sac), but the clinical appearance did not show any change. She underwent several incisions whenever the abscess severely exacerbated, threatening to perforate spontaneously. Attempts to probe the nasolacrimal canal were performed in the “silent” phases of the disease, but the probe did not pass deeper than the bottom of the lacrimal sac, suggesting the absence of the canal. Attempts to irrigate the lacrimal canals were not successful either. A lacrimal sac massage resulted in a certain amount of mucus bursting out from the inferior lacrimal punctum, indicating a blockage of the lacrimal system underneath the lacrimal sac.
Endonasal endoscopic examination showed no signs of intranasal extension.
Because of the clear clinical diagnosis of dacryocystocoele, the patient's age, and the need for additional general anaesthesia, we did not insist on a dacryocystogram or computed tomograph scanning.
At the time we decided to try to perform an endonasal endoscopic dacryocystorhinostomy (EEDCR), the girl was in one of her “silent phases”, without any clinical sign of acute exacerbation of the infection. Only moderate hemispheric bulging was seen in the medial canthal region.
We started the procedure by inserting two small, very thin (20×5 mm) gauze flakes, previously soaked in a 5% cocaine solution mixed with adrenaline (5:1 ratio) and then firmly squeezed, into the left nasal cavity. The flakes were removed after 5 minutes. A favourable vasoconstriction of the whole nasal mucosa was achieved. Then 0.5 ml of local anaesthetic (1% lignocaine with 1:100 000 adrenaline) was injected submucosally in the area just anterosuperior to the insertion of the middle turbinate. We used a paediatric endoscope of 2.7 mm in diameter and 30 degrees optics. The mucosa of this region was then removed by means of bipolar coagulation, and lacrimal bone was nicely exposed in an oval shape measuring up to 6 mm in longer diameter. The bone was drilled off and thinned out, so that the lacrimal sac became visible (the removal of the underlying lacrimal bone is more easily performed posteriorly, where it is thinner, but is more safely performed anteriorly to avoid the possibility of orbital disruption). The ophthalmological probe was inserted into the lacrimal canal and the sac itself, tenting it towards the nasal cavity. Then lacrimal sac marsupialisation was performed, using otological microsurgical scissors and punches, and a large amount of turbid tears mixed with mucopurulent discharge was obtained.
A 6 cm long nasal thin gauze ribbon package with antibiotic ointment was placed in the operated region for 3 days. After 5 days, there was no sign of dacryocystocoele or dacryocystitis on the girl's face (Fig2). Eight months after the surgery, she is feeling fine.
Since, after 16 weeks of life, the nasolacrimal duct obstruction and dacryocystocoele did not resolve spontaneously or after conservative treatment, we performed an endonasal endoscopic dacryocystorhinostomy. As far as we know, this was the youngest child ever operated by means of EEDCR. The small anatomical dimensions of the infant nose posed a technical challenge in performing EEDCR: during the use of Richard's otological drill for bone removal, there was some difficulty in concomitant endoscopic visualisation and potential damage to the nasal mucosa from rotation of the drill shaft. We also performed a lacrimal opening of 6 mm with angled endoscopic biting forceps (the usual opening is about 10 mm).
In comparison with an external dacryocystorhinostomy, EEDCR avoids an external scar and offers very low morbidity in the immediate postoperative course. In spite of the technical problems, we think that EEDCR, even in such a small infant, can be a good therapeutic choice in cases refractory to conservative treatment (warm compresses, massage, probing) because of its non-invasive performance and a very fast postoperative rehabilitation.