Discussion

The lacrimal apparatus is in close proximity to the uncinate process in the lateral wall of the nasal cavity. As uncinectomy is part of FESS, there is a chance that the nasolacrimal duct will become injured during FESS. The following anatomical points regarding the nasolacrimal duct deserve due consideration while performing uncinectomy in FESS.^{Reference Song, Jin, Kim, Huh, Kim and Kim8–Reference Zhang, Cheng, Xie, Wang, Zhang and Chen12}

The lacrimal bone is situated immediately anterior to the mid-third of the uncinate process, which makes it vulnerable to injury during uncinectomy.^{Reference Chastain, Cooper and Sindwani13}

The nasolacrimal sac is only 1–8 mm anterior to the root of the uncinate process and 0.5–1.8 mm from the natural ostium of the maxillary sinus. This makes it vulnerable to injury, not only during uncinectomy but also during the widening of the maxillary sinus ostium in FESS.^{Reference Cohen, Antunes and Morgenstern5}

There is a dehiscence of lacrimal bone in 20 per cent of the normal population.^{Reference Unlu, Goktan, Aslan and Tarhan3} The lacrimal bone is very thin (0.1 mm): its thickness varies from between 100 μm and 300 μm, so it can be easily penetrated by surgical instruments.^{Reference Hartikainen, Aho, Seppa and Grenman14} The bony wall separating the mucous membrane of ethmoidal cells from the lacrimal bone is often missing.^{Reference Bolger, Parsons, Mair and Kuhn7}

In the case of extensive pneumatisation, there may be pneumatisation of agger nasi cells (with anterior extension), the lacrimal bone and the frontal process of the maxilla. This leads to thinning of the bone over the lacrimal sac and duct, posing a risk of injury.^{Reference Bolger, Parsons, Mair and Kuhn7,Reference Whitnall15}

It would also be pertinent to note that the bony nasolacrimal canal is not delineated during nasal endoscopy. The bony nasolacrimal canal is highly variable in size, with differences associated with age, sex and race. Moreover, the lacrimal bone may be subject to patchy erosion associated with ageing.^{Reference Shigeta, Takegoshi and Kikuchi16}

All these factors make the nasolacrimal duct prone to injury during FESS. This also underlines the importance of the surgical technique that needs to be adopted during uncinectomy, so as to minimise the said injury.

We recorded a pre-operative dehiscence of 1.16 per cent in this study. This is markedly lower than the 20 per cent quoted by Unlu et al.^{Reference Unlu, Goktan, Aslan and Tarhan3} and 6.8 per cent reported by Ali et al.^{Reference Ali, Murphy, Wormald and Psaltis2} This reflects the prevalence of nasolacrimal duct dehiscence in the subcontinent and adds to the minimal data on this subject.

In the present study, only one case of nasolacrimal duct injury was recorded following FESS in 43 patients (86 surgical procedures). This makes the injury rate 1.16 per cent. The review of literature on this subject reveals an injury rate of up to 15 per cent for nasolacrimal duct injury in patients undergoing FESS.^{Reference Bolger, Parsons, Mair and Kuhn7} However, only 0.3–1.7 per cent of these cases were symptomatic.^{Reference Serdahl, Berris and Chole6,Reference Bolger, Parsons, Mair and Kuhn7,Reference Kennedy, Zinreich, Shaalan, Kuhn, Naclerio and Loch17,Reference Hunink, de Vries-Knoppert, Balm and Luth18} Our impressive results could be attributed to the surgical expertise of the principal investigator (GBS), an experienced rhinologist, and the cautious use of backbiting forceps while performing ‘swing door’ uncinectomy.

Interestingly, only a few studies have evaluated nasolacrimal duct injury in FESS. The salient features of these studies are given in Table 1. Kennedy et al. (1987) for the first time reported 2 symptomatic cases of nasolacrimal duct injury while performing endoscopic middle meatus antrostomy in 117 surgical procedures.^{Reference Kennedy, Zinreich, Shaalan, Kuhn, Naclerio and Loch17} Later, Serdahl et al. (1990) reported the treatment of eight cases of nasolacrimal duct injury by dacryocystorhinostomy.^{Reference Serdahl, Berris and Chole6} Serdahl, an ophthalmologist, analysed these cases by Jones I and Jones II tests. All these cases were referred to him, and thus the exact sample size is not available. Thereafter, Bolger et al. (1992) reported 7 cases of nasolacrimal duct injury in 46 FESS operations (sample size of 24 patients).^{Reference Bolger, Parsons, Mair and Kuhn7} This was a prospective study and used fluorescein dye to detect the injury. Later, Unlu et al. (2001), in a prospective study comprising 31 patients, reported 53.2 per cent dehiscence using active transport dacryocystography. None of their patients were symptomatic.^{Reference Unlu, Goktan, Aslan and Tarhan3} Nasolacrimal duct injury has also been reported by Saengpanich et al. (2001) in microscopic sinus surgery.^{Reference Saengpanich, Kerekhanjanarong, Chochaipanichnon and Supiyaphun19} In that study, only 1 out of 32 patients had nasolacrimal duct injury, and that too was asymptomatic. It is important to note that none of these studies used radiology for the evaluation of the nasolacrimal duct.

Table 1. Synopsis: studies on nasolacrimal duct injury following FESS

FESS = functional endoscopic sinus surgery; NLD = nasolacrimal duct; N/A = not applicable

Radiological evaluation of the nasolacrimal duct has been reported in the medical literature.^{Reference Takahashi, Kakizaki and Nakano20–Reference Aggarwal, Sharma, Das and Singh23} However, in an extensive internet search using PubMed and Medline services, we could find only one study on the radiological evaluation of nasolacrimal duct injury after FESS. In this retrospective study by Ali et al., comprising 63 patients, a pre-operative nasolacrimal duct dehiscence rate of 6.8 per cent and a post-operative dehiscence rate of 3.3 per cent (4 patients) were recorded for 118 surgical procedures.^{Reference Ali, Murphy, Wormald and Psaltis2} Out of these four patients, two had symptomatic epiphora, one of which required dacryocystorhinostomy. One of the important drawbacks of this study is the non-interpretation of nasolacrimal duct injury by the radiologists. In contrast, the present study is prospective in nature with the active involvement of a radiologist who is well conversant with the reporting of nasolacrimal duct dehiscence and lamina papyracea anatomy. This is in the best interests of patients and would also act as a safeguard against medicolegal litigation.

The technique of conventional cannulation dacryocystography is invasive, requiring local anaesthesia, cannulation, dilatation of the lacrimal punctum and injection of contrast material. There is also a risk of iatrogenic trauma or scarring of the lacrimal apparatus. The technique also requires co-operation from the patient.^{Reference Zinreich, Miller, Freeman, Glorioso and Rosenbaum24,Reference Weber, Rodriguez-DeVelasquez, Lucarelli and Cheng25} In contrast, radiological evaluation of the nasolacrimal duct is a simple and non-invasive method. Moreover, pre-operative dehiscence detected in this way guides the surgeon to carry out uncinectomy with caution, so as to prevent inadvertent injury to the nasolacrimal duct.

Modern radiological evaluation of the nasolacrimal duct apparatus involves the use of topical instillation of contrast material or saline with CT or magnetic resonance imaging (MRI).^{Reference Caldemeyer, Stockberger and Broderick26,Reference Rubin, Bilyk, Shore, Sutula and Cheng27} We believe that this technique is more beneficial in cases of epiphora, as it evaluates the physiology of the nasolacrimal duct apparatus. Since 1990, MRI has been used as an adjunctive diagnostic test for evaluating the nasolacrimal system. When combined with a contrast agent, it allows very fine resolution of soft tissue structures within and surrounding the nasolacrimal system.^{Reference Goldberg, Heinz and Chiu28} However, it is expensive, it has poor ability to image bony structures (not conducive to FESS), and there can be artefacts from the nearby ethmoidal cells. Magnetic resonance imaging is also susceptible to movement artefacts because of the relatively long acquisition times required.^{Reference Dutton, White, Cohen, Mercandetti and Brazzo29} Moreover, magnetic resonance dacryocystography visualises the smaller draining structures inconsistently.^{Reference Caldemeyer, Stockberger and Broderick26} Hence, it is not currently recommended for routine use.^{Reference Dutton, White, Cohen, Mercandetti and Brazzo29}

Some may find the irradiation dose absorbed by the eye lens in spiral CT to be high (24.5 mGy, as compared with 1.2 mGy for digital subtraction dacryocystography), but it is still well within the prescribed radiological clinical limits (permissible dose: 0.5–2 Gy).^{Reference Zammit-Maempel, Chadwick and Willis30} Technically, the recommended radiation dose to the lens is 150 mSv annually. A CT of the paranasal sinuses generates only 5 mSv of radiation; thus, multiple scans pose no radiation hazard to the patients and can be conducted safely.^{Reference Rehani, Vano, Ciraj-Bjelac and Kleimen31}

There are caveats to this study. The study cohort represents a population referred to a single tertiary care teaching hospital. Thus, the data recorded may not be representative of the general population. In addition, this study was not randomised and the results were not ascertained blindly. Observational studies such as these are also subject to potential biases (e.g. selection bias) and confounding. Further, the results might have been influenced by the expertise of the operating surgeon (GBS), a senior faculty member with considerable experience in FESS. An additional study limitation is the sole employment of ‘swing door’ uncinectomy. Last but not least, the study fails to compare the radiological technique versus conventional dacryocystography in the assessment of nasolacrimal duct injury.

The true strength of this study lies in its prospective nature. This eliminates the shortcomings of a retrospective study: incomplete and inaccurate clinical data, and a poor follow-up record. This study amalgamates the realities of clinical practice and the rigours of scientific analysis, and thus further adds to the sparse medical literature on this subject. The true value of this study in the context of existing literature is the evaluation of nasolacrimal duct injury by non-invasive radiology in a prospective study design, hitherto unreported.