Endoscopic sinus surgery has been widely used for most sinonasal disease including that in the maxillary sinus. There are various lesions that can be present in the maxillary sinus such as chronic inflammation, odontogenic infection, and neoplastic disease. In cases with sinus inflammation or infection are controlled by functional endoscopic sinus surgery to reestablish ventilation and mucociliary clearance of the sinuses [1]. However, in cases of benign sinonasal tumor such as inverted papilloma, resection of its attachment site should be performed through elevation of a mucoperiosteal flap surrounding the lesion with safe macroscopic margins. Incomplete removal can lead to early recurrence of the tumor [2]. It is common to perform middle meatal antrostomy first to remove maxillary sinus-occupying lesions. Considering the position of the maxillary sinus although we can visualize lesions in the maxillary sinus with a 70º or 90º endoscope, it is often difficult to remove the lesion completely because the limited reach of the instrument. In particular, when the lesion is located on the inferior or anterior side of the maxillary sinus, it usually hard to remove through the natural ostium. Such removal can be performed using the traditional Caldwell-Luc approach (CLA) or by way of a canine fossa puncture [3,4]. Complications of the canine fossa approach have occurred in up to 75% of patients and include cheek swelling, dental numbness, facial numbness, tingling, and pain [5,6].

Zhou et al. [7,8] proposed the intranasal endoscopic prelacrimal recess approach to the maxillary sinus. This procedure is performed by removal of the medial wall of the prelacrimal recess, preserving the inferior turbinate and nasolacrimal duct. Many cases have been reported using this approach to remove unilateral maxillary sinus diseases such as inverted papilloma and chronic inflammation [9], tumor of pterygopalatine fossa [10], and tumor located at the infratemporal fossa [11]. Endoscopic prelacrimal medial maxillectomy (EPMM) has the advantage of avoiding an external incisional wound.

However, many surgeons are reluctant to use this approach because they consider its limitations relative to CLA. Comparison between CLA and EPMM is also not well established. This study compared EPMM with conventional CLA to remove benign maxillary sinus tumor and to evaluate the usefulness of these approaches based on our experience.

This retrospective study was approved by the Institutional Review Board of Samsung Medical Center (IRB No. 2018-05-129). The informed consent was exempted because the risk to the study subjects was extremely low.

Demographic data of benign maxillary sinus tumor patients in this study is presented in Table 1. A total of 40 patients (32 males and eight females) diagnosed with maxillary sinus-occupying benign tumor by radiological imaging modality were enrolled. We divided the study population according to surgical approach–CLA and EPMM. The mean age of the CLA group was 53.7 years, and that of the EPMM group was 53.6 years. The proportion of previous operation history was 40% in both groups. Mean age (P=0.978) and follow-up period (P=0.618) between the two groups were not significantly different. Pathological results showed that all 30 CLA group cases were inverted papilloma, and EPMM comprised eight inverted papilloma, one ameloblastoma, and one ossifying fibroma case.

Comparison of surgical outcomes according to surgical approach is presented in Table 2. There were no cases of failure of gross total removal during surgery, and no recurrences were observed during the follow-up period for either group. The operation time (P=0.825), bleeding amount (P=0.999), and hospital stay (P=0.397) were also not significant between two approaches. Squamous cell carcinoma was seen in one patient in the EPMM group. This patient underwent additional postoperative radiation therapy and had regular follow-up at outpatient clinic. Eleven CLA patients (37%) and three EPMM patients (30%) had numbness around the cheek and upper lip area (P=0.715). Two CLA patients had facial pain after surgery. Epiphora was not been observed in the EPMM group. In the CLA group, there were eight patients who had numbness lasting more than 3 months and two patients for more than 1 year. However, facial numbness disappeared within 3 months in all EPMM patients.

The attachment site of the benign maxillary sinus tumor is shown in Table 3. Evaluation of attachment site was performed through surgical procedures. In the CLA group, the attachments in 30 cases of benign maxillary sinus tumor were found at the anterior wall (n=8), posterior wall (n=2), lateral wall (n=2), medial wall (n=6), superior wall (n=5), inferior wall (n=4), and diffuse attachment to the maxillary sinus (n=3). In the EPMM group, the attachments in 10 cases of benign maxillary sinus tumor were found at the anterior wall (n=3), lateral wall (n=1), medial wall (n=2), inferior wall (n=1), and diffuse type (n=3). Fig. 2 shows three cases using EPMM for maxillary sinus inverted papilloma. Tumors were successfully removed. Fig. 3 shows another case using EPMM. A 48-year-old female patient visited for left cheek tingling sensation of 4-year duration. CT scan showed a fibro-osseous lesion encasing the infraorbital nerve at the maxillary sinus anterior wall. We performed tumor removal and infraorbital nerve decompression through EPMM, and her symptoms resolved after surgery. Ameloblastoma removal via EPMM is presented in Fig. 4.

Endoscopic sinus surgery is frequently performed for sinonasal disease, and middle meatal antrostomy is the most basic technique in the surgical procedure. We focused on benign maxillary sinus tumors in using EPMM as originally described by Zhou et al. [7]. In many cases of inflammation or infectious disease such as odontogenic sinusitis, fungal sinusitis, chronic sinusitis, and nasal polyp, we do not usually remove attached mucosa with a free margin. Because the purpose of surgery in these diseases is to reestablish ventilation, mucociliary clearance of the sinuses with removal of irreversibly changed tissue. Therefore, in assessing whether it is a suitable surgical procedure for maxillary sinus disease, it would be better to confine the study population to tumorous lesions. We also excluded cases of malignancy because an external approach should be considered when there is bony invasion. Finally, we reviewed 40 patients who needed an additional surgical approach for benign maxillary sinus tumor removal and among these patients; 10 patients underwent surgery through EPMM.

There are many types of benign tumor originating from the sinonasal cavity including inverted papilloma, hemangioma, fibrous dysplasia, and angiofibroma. Of these, inverted papilloma, which accounted for most of our study population, has a high recurrence rate and a chance of malignant transformation. These tumors are also well known to exhibit centrifugal growth pattern [12,13]. Even though such a tumor usually occupies the nasal cavity with large volume, the origin is often located in a focal area. Therefore, it is important to predict the attachment site of the tumor through preoperative radiological imaging, and the tumor must be completely removed around the attachment site. Lee et al. [14] reported that focal hyperostosis on CT could be considered a predictor of inverted papilloma origin. In our study population, 40% of all patients had a previous operation history at another hospital and visited our center for recurred lesions. In most cases, only middle meatal antrostomy was performed as primary surgery. Therefore, the lesions could not be removed from the attached origin site properly. There was no difference in proportion of patients with previous operation history between the two groups.

The maxillary sinus is a pyramidal shaped chambered space that is usually the largest sinus of the paranasal space, and its natural ostium is formed at the superior and medial areas of the maxillary sinus. Because of these anatomical characteristics, the inferior and anterior walls of the maxillary sinus are difficult to approach despite using angled microdebrider and endoscopy. It has been reported that even the experienced sinus surgeon can have trouble accessing all walls of the maxillary sinus [15,16]. Historically, the difficult-to-approach area of the maxillary sinus was accessed via the CLA [17].

In 2007, Zhou et al. [8] first reported performing operation using EPMM for maxillary sinus disease. Since then, many surgeons have reported cases using this approach. However, comparison between CLA and EPMM has not been presented. The indications for and limitations of this approach are also not well established. In our experience, there are two important advantages in using EPMM. One is to avoid nasolacrimal duct injury when opening a lacrimal bone surrounding a nasolacrimal duct. Second point is to remove the medial wall of the prelacrimal recess widely to facilitate instrument use. EPMM is basically performed by removing the maxillary sinus medial wall of the prelacrimal recess, which is the bone between the pyriform aperture and nasolacrimal duct. However, after full exposure of the nasolacrimal duct and medialization with lateral nasal wall mucosa can be done, we perform bony removal of maxillary sinus medial wall to the natural ostium, which is the anterior margin of the middle meatal antrostomy. This allows a wide surgical field, making it easier to use the instruments. We used these tips in some cases with widely attached tumor of the maxillary sinus.

In our study, three patients of the EPMM group (30%) had facial numbness around the cheek and upper lip area after surgery. Although our study showed no significant difference in incidence of facial numbness between the two groups, the duration of numbness was significantly shorter with EPMM than with CLA. These neurologic complication is thought to be caused by injuries to the branches of the infraorbital nerve, principally the anterior superior alveolar nerve and less commonly the middle superior alveolar nerve [3]. When drilling the pyriform aperture, we can damage the nerve. However, nerve transection is merely happened during the EPMM and transient numbness could be happened because of thermal injury by cauterization or drilling. However, when making antral window in CLA, we can make damage to the main branch of infraorbital nerve. When elevating periosteum up to the infraorbital canal, infraorbital nerve could be in risk. If the infraorbital artery injured, cauterization could also make permanent damage to infraorbital nerve.

Even when compared with endoscopic medial maxillectomy as a transnasal surgical approach, EPMM has an advantage of reducing complications by preserving the inferior turbinate and lacrimal duct. Osguthorpe and Weisman [18] performed medial maxillectomy in 35 patients and reported that all patients complained of nasal crusting, which is a symptom of empty nose syndrome. Another study reported a 30% incidence of epiphora after medial maxillectomy [19]. Our patient did not complain of persistent nasal crusting or epiphora. In our EPMM cases, we performed one or two sutures at incision site using 4-0 vicryl and there was no adverse event such as necrosis or detachment of interior turbinate. During the follow-up period, it was difficult to exam the tumor recurrence through the inferior meatus in EPMM because the lateral nasal wall mucosa of inferior meatus was preserved. However, because all the EPMM cases were performed after middle meatal antrostomy was conducted in advance, we were able to exam the recurrence of tumor through the middle meatal antrostomy site using 30º, 70º, and 90º endoscopy.

The main advantages of EPMM is a wide surgical field for all maxillary sinus walls without violation of the nasolacrimal duct and inferior turbinate. Our study showed that benign maxillary sinus tumors attached to various sites of the maxillary sinus could be removed, and gross total resection is possible even for diffuse-type tumors with this approach. In addition, since EPMM is performed via a transnasal method, it can preserve the periosteum of the canine fossa area, which is the manipulated site of the CLA. Periosteum plays a role as a primary barrier to prevent maxillary sinus disease from invading the skin of the cheek. When CLA is performed for maxillary sinus inverted papilloma, if the final pathology result is confirmed as squamous cell carcinoma, the approach site could become a spreading route. Therefore, EPMM has an advantage that the periosteum of the canine fossa area can be left as a barrier. There are some limitations of EPMM. Beyond the standard complications of endoscopic sinus surgery, potential complications of EPMM include damage to the nasolacrimal apparatus, nasal dryness, and facial numbness [20]. Additionally, in patients with a narrow nasal cavity due to a septal deviation or maxillary sinus hypoaeration, it can be difficult to perform EPMM based on our experience.

If there is no reluctance to sublabial incision, both approaches can be used for all maxillary benign tumor according to surgeon’s preference. However, based on our experiences, we would like to discuss the strengths and weaknesses of EPMM and CLA. Anterior wall of maxillary sinus tumor is more efficiently managed by EPMM than CLA. On the contrary, superomedial wall is relatively difficult to be manipulated by EPMM compared with CLA. Although anteroinferior wall is also difficult to approach by both approaches, there is more chance to damage of anterior inferior alveolar nerve when drilling anteriorly in performing EPMM.

This study has some limitations. First, the sample size of the EPMM group was too small to establish the effectiveness of this approach. Second, the range of maxillary sinus disease between the two groups was not the same. However, this study showed that all walls of the maxillary sinus can be accessed and total gross removal of tumor can be performed through EPMM. The follow-up period is quite short in our study. Because they were mostly asymptomatic patients, the rate of follow-up over 1 year was not high. Although further study of a large population is needed to establish definite indications and limitations of EPMM, our study offers a comparison between the two approaches, showing that tumor control is comparable. EPMM is also a useful method for diverse benign sinonasal tumors in maxillary sinus.

EPMM is a useful approach for benign maxillary sinus tumors. This approach can be used for benign tumors of the entire wall of the maxillary sinus without lacrimal duct or inferior turbinate injury. Tumor control is comparable with that of CLA. The duration of facial numbness after surgery was also significantly shorter than that with CLA.