The positional relationship between the lacrimal sac and skull base was examined in this study. On the anterior plane, the frontal sinus lies between the lacrimal sac and skull base in most cases, which prevents transmission of twisting and rotational forces to a rongeur and following indirect injury to the skull base. The skull base and top of the frontal sinus were considerably found to be away from the lacrimal sac fundus. Although the maxillary bone around the anterior lacrimal crest is considerably thick (the maximum thickness, 4.6–6.3 mm) [14, 15], the thickness of the skull base was similar. Furthermore, the frontal bone at the level of the top of the frontal sinus was thicker than the maxillary bone. The posterior plane was farther from the lacrimal sac fundus. The cribriform plate was positioned above the horizontal level of the lacrimal sac fundus in most cases. These results indicate that the risk of CSF leakage by both direct and indirect injuries to the skull base during the creation of a bony window in DCR is extremely low.
Primary nasolacrimal duct obstruction and dacryocystitis are more common in females , and they usually have a small face, which had proposed an expectation that the distance from the lacrimal sac fundus to the skull base is shorter in females. However, our study showed that there was no significant sex-related difference in the distance from the lacrimal sac fundus to the skull base, indicating that the risk of CSF leakage is not different between sexes.
Our study demonstrated that 9.9–18.2% of cases had no frontal sinus interposition. In such cases, the cribriform plate was anteroposteriorly and horizontally closer to the lacrimal sac fundus than those with frontal sinus interposition. In this situation, a superior and posterior extension of the creation of a bony window may cause direct injury to the skull base. However, the skull base was considerably far from the lacrimal sac fundus and was significantly thicker in cases without frontal sinus interposition. Furthermore, the cribriform plate formed the inferior boundary of the olfactory fossa 7.6 mm posterior to the posterior lacrimal crest. Therefore, direct injury of the skull base/cribriform plate is rarely expected. On the contrary, the twisting movement of a bone rongeur during bone removal can be transmitted to the skull base/cribriform plate in cases without frontal sinus interposition. Surgeons should avoid the twisting and rotational movements in cases without frontal sinus interposition to prevent indirect injuries to the skull base/cribriform plate.
Among patients with frontal sinus interposition, the cribriform plate was positioned below the horizontal level of the lacrimal sac fundus in 13.7% of patients. In such patients, the skull base was also closer to the horizontal level of the lacrimal sac fundus on the posterior plane. However, the distance from the posterior lacrimal crest to the posterior plane was 17.9 mm, indicating that direct injury of the skull base/cribriform plate is improbable.
Previous articles recommended using the MCT as a landmark for the creation of a bony window without skull base injury [3, 10, 12, 17, 18]. Our previous study demonstrated that among 75 orbits from 48 cadavers, the maximum distance between the CCO and superior edge of the MCT was 2.32 mm . This result was similar to that in our present study that the maximum distance between the CCO and superior edge of the MCT was 2.58 mm. As complete exposure of the CCO is an important step of DCR to increase the success rate of DCR , these results imply that the bone removal till the level of around 3 mm above the MCT is necessary for complete exposure of the CCO. On the contrary, Kurihashi et al. found that on the coronal section through 10 mm posterior to the medial canthus, the skull base was positioned only 3 mm above the horizontal level of the medial canthus in 21.4% of cadavers . Based on this result, several articles advocated that a bony window should not be extended 3 mm above the MCT [3, 17, 18]. Our present study also found that the cribriform plate was positioned below the horizontal level of the lacrimal sac fundus in several cases, and the maximum negative vertical distance was − 3.52 mm. However, as a matter of course, the CCO was always located below the lacrimal sac fundus, and the mean distance from the CCO to the lacrimal sac fundus was 5.3 mm, indicating that when bone removal extends superiorly till the exposure of the CCO, the superior margin of the bony window does not usually reach the horizontal level of the cribriform plate.
The cribriform plate is thought to be located more inferiorly in Asians . The study by Neuhaus et al. in Caucasians demonstrated that the mean vertical distance from the superior margin of a bony window to the cribriform plate at the level of the posterior limit of the frontal sinus was 5.7 mm . Similarly, Botek et al. showed that the mean vertical distance from the CCO to the most anteroinferior part of the cribriform plate was 15.13 mm . In our study, the mean vertical distance of the horizontal level of the lacrimal sac fundus from the cribriform plate was 4.2 mm. Although the reference points for measurements were different between these previous [1, 11] and our present studies, the Japanese seem to have a lower-positioned cribriform plate.
To avoid twisting and rotational forces to the rongeur, alternative use of a drill and an ultrasonic bone aspirator is useful [19, 20]. For patients with a high risk of CSF leakage during DCR, such as pre-existing deformity and an anatomical abnormality of the skull base [5, 6], the use of an image-guided navigation system may be helpful to avoid this complication .
If CSF leakage occurs during DCR, the location and size of the dural injury should be confirmed . If the leak is small, this may be spontaneously closed with the patient’s head kept elevated . Fat tamponade, tissue fibrin glue, and bone wax are used to close the point of leakage . Patients should be monitored postoperatively, and consultation with neurosurgeons and otorhinolaryngologists should be obtained as appropriate .
There are a few limitations to this study. First, this study included only Japanese cadavers and patients. Since there are known racial differences in lacrimal and skull base anatomy , the findings presented here may not be applicable to other races. Second, all the measurements were performed by a single examiner, which could affect the reliability of this study. Third, direct measurement of the anteroposterior distance in cadavers may be more precise.
In conclusion, the positional relationship between the lacrimal sac and skull base was examined in this study. The results of our study indicate that the risk of CSF leakage during DCR is extremely low because of a high rate of the interposition of the frontal and ethmoid sinuses, the lacrimal sac positioned far away from the skull base/cribriform plate, and a thick skull base. However, oculoplastic surgeons should be cautious of this complication by indirect injury in patients with no interposition of the frontal and ethmoid sinuses. It is essential to confirm the existence of sinus interposition on CT images before performing DCR.