Endoscope-assisted resection of nonneoplastic space-occupying lesion in oral and maxillofacial areas

Endoscope-assisted oral and maxillofacial surgeries have been applied to the resection of tumors with minimal invasion and good cosmetic outcomes. However, with regard to endoscope-assisted resection of nonneoplastic space-occupying lesion (NSOL) in oral and maxillofacial areas which differ from tumors in treatment, there are no systematic reports. Therefore the advantages and limitations of the endoscopy-assisted approach (EAA) in resection of NSOL remain unclear. In this novel study we describe endoscope technique for resection of NSOL in face and submandibular areas and compare the feasibility and effectiveness of EAA with external approach (EA). Eleven patients underwent EAA and 20 patients underwent EA procedures. The perioperative and postoperative outcomes of the patients were evaluated. The resection of NSOL with EAA was completed successfully with a shorter hospitalization duration, less bleeding, a smaller incison and better satisfaction with appearance than with the EA procedure (P < 0.01). Our study showed that endoscope-assisted resection of NSOL is technically safe, feasible and practicable. Good cosmetic results with minimal invasion can be achieved with this new technique and therefore this may be a promising new standard procedure in oral and maxillofacial areas.


Materials and Methods
Subjects. Thirty-one patients with different NSOL underwent surgery between November 2012 and April 2016 in the Department of Oral and Maxillofacial surgery, at the Second Hospital of Shandong University. There were 11 male patients and 20 female patients, with ages ranging from 16 to 68 and a median age of 37 years. EAA were performed on 11 patients with NSOL in oral and maxillofacial areas, of which two patients were diagnosed as submandibular epidermoid cyst, two patients were diagnosed as IVM in front of parotid gland and seven patients suffered from submaxillaritis. The remaining twenty patients were treated with EA and included 5 submandibular epidermoid cysts, 3 IVM and 12 submaxillaritis. Computed tomography (CT) scan and/or magnetic resonance imaging (MRI) were used to evaluate the lesions of all the patients prior to surgery. All neoplastic space-occupying lesion or suspected cases of gland tumors were excluded from this study. This study was approved by the Ethics Committee of the Second Hospital of Shandong University. All patients received detailed information about the operative approach and signed informed consent prior to participating in this study. The patients with EAA were all informed that a conventional wide-open operation may be required if any surgical complications were encountered e.g. uncontrolled bleeding that could not be resolved by endoscope-assisted surgery, and that all the excised samples would be diagnosed by fast frozen pathology. If the patient was diagnosed with a tumor, a conventional open procedure would also be carried out. Statistical analysis of all of the data was performed using SPSS for Windows (SPSS Inc., Chicago, IL). Data are presented as mean values ±SD. For all analyses, the statistical differences were considered to be significant if P < 0.05. Surgical procedure. Under general anesthesia, the patient's neck was placed in the supine position with a pillow under the shoulder and extended, the head was then rotated to the opposite side of the lesion. A video camera system (Karl Stortz, Germany), a 30° 4-mm endoscope and a 0° 4-mm endoscope (Karl Stortz, Germany) were used (Fig. 1A,B). A 15-to 30-mm skin incision was made between the inferior and superior margin of the submandibular lesion, preferably in a natural cervical wrinkle over the middle of the protruding dome of the lesion ( Fig. 2A). The CT examination results helped operators to evaluate the scope of the cyst in advance (Fig. 2B-D). The lesions (glands or cysts) were exposed after the incision of skin, subcutaneous tissue and platysma muscle. Then the dissection proceeded along an avascular plane round the pseudocapsule between the lesion and the adjacent tissue. This can decrease bleeding and get clear fields while minimizing the possibility of damaging the facial artery, vein and the marginal mandibular branch of the facial nerve. Special retractors, bipolar coagulation forceps and ultrasonic scalpels were also used in order to decrease the amount of bleeding and obtain clear fields (Fig. 2E). The first operation assistant was responsible for the correct positioning of the endoscope during the procedure. The main task of the second assistant was to provide maximum working space by lifting the skin flap away with two retractors. The facial artery and vein, Warthon's duct and loop of the lingual nerve were identified and suture-ligated for the patients with submaxillaritis. As for epidermoid cysts the dissection along an avascular plane around the cyst was performed (Fig. 2F) and the cysts were removed through the surgical wound (Fig. 2G). The interior wounds were closed in layers with 5-0 absorbable sutures after irrigation, and the skin was stitched with 5-0 nylon sutures. A silastic drain was inserted deep in the wound and kept in situ for at least 2 days after the surgery (Fig. 2H).
In surgeries of the lesion in front of the parotid gland the same anesthesia and positioning of instruments were performed, with two assistants assisting the operator by positioning the endoscope and providing working space. The location of the lesion was marked on the face pre-operatively (Fig. 3A). A small tragus incision 2 cm to 2.5 cm long was made and did not extend beyond the inferior margin of the earlobe (Fig. 3B). An under skin tunnel was formed along the avascular plane around the surface of parotid gland capsule. The tunnel was extended to the mass with the same plane around the lesion, which was clearly observed under endoscopic visualization (Fig. 3C). The lesion was resected ( Fig. 3D) with careful protection of nerves and other normal anatomic structures. Negative pressure drainage was removed two days after the operation (Fig. 3E).   recorded every 24 hours, and the silastic tubes were removed once the drainage had reduced to less than 10 mL in a 24-hour period. The operative incision was defined as infection if it appeared obviously red and swollen with pus or required antibiotic treatment 27 . Postoperative pain one week after the operation was evaluated based on the visual analogue scale [27][28][29] . Patient satisfaction with cosmetic appearance at 3 weeks after surgery was also assessed using a visual analogue scale. The median follow-up durations ranged from 9-24 months after leaving hospital.

Results
Clinical data of the patients and statistical analyses of all of the data are shown in Tables 1 and 2. All of the lesions in the 31 patients were resected completely and no patient suffered from complications such as infection, nerve injury or excessive bleeding. All eleven patients with EAA diagnosed by intraoperative frozen sections were treated by resection with the endoscopically assisted system and without having to revert to EA. There were significant statistical differences in the length of the wounds and the amount of intraoperative bleeding between the patients treated with EAA and those with EA (P < 0.001). However, the EAA procedure has a longer operation time compared with the EA procedure (90.64 ± 20.87 vs 52.05 ± 6.82 min, P < 0.001). Comparison of postoperative drainage of the two methods showed no statistical difference, lso there was no significant difference in the postoperative pain score (P = 0.082). The average hospital stay of the patients with the EAA were shorter compared with the EA group (2.91 ± 0.74 vs 3.78 ± 0.70 day, P = 0.003). All the patients with EAA were significantly more satisfied with their cosmetic outcomes than those with the EA (P < 0.001). There were no postoperative complications, including recurrence, chronic infection, Frey syndrome or pain in any of the patients followed up for 9-24 months after leaving hospital and good cosmetic results were achieved in the long term follow up (Fig. 4A,B).

Discussion
Endoscope use was initially introduced to oral and maxillofacial surgery in laser lithotripsy of salivary gland stones by Königsberger R et al. 1 . Since then many surgeons have reported the application of endoscope use in oral and maxillofacial areas as it has several advantages over the conventional approach, including reduced tissue damage, a smaller wound, fewer wound-related complications, and minimal postoperative scarring 11,30-33 . Among the many advantages the good cosmetic result is the most satisfactory outcome to both patients and surgeons 14 than in laparoscopic surgery, thoracoscopic surgery and endoscopic sinus surgery there was no natural anatomical space for endoscope-assisted surgery in oral and maxillofacial areas. Therefore the working space was formed artificially and was poor compared to natural space 14 . There is a general consensus that the tumors from parotid or sunmandibular gland should be partially excised around extracapsular tissue [18][19][20] . Given the poor operation space endoscope-assisted partial resection around extracapsular tissue requires a highly skilled technique, which prevents its popularization. Therefore, such operations are not yet a standard procedure in the head and neck region 14,36 In this study, endoscope-assisted surgeries were performed on patients with NSOL in oral and maxillofacial areas with less intraoperative blood loss, shorter hospital stays and better cosmetic outcomes than those with EA. Subcapsular dissection was sufficient for radical treatment of all the patients without concerns about the incompletion of pseudocapsule and satellitosis of the tumor. All the endoscope-assisted surgeries were completed successfully. All patients received satisfactory cosmetic outcomes and no postoperative complications occurred during the months following surgery. In comparison resection of the submandibular lesions with EA is performed through a long cervical incision, which can result in a visible and troublesome scar. Additionally, the chance of damaging the marginal mandibular branch of facial nerve has been reported to be 1-7% in patients 37,38 . Endoscope-assisted technique allows for the manipulation of tissues in small spaces and provides improved access to lesions that might not be attainable by conventional surgical procedure. It provides excellent surgical exposure and achieves complete resection of the tumor with good haemostasis and minimum morbidity, while preserving the key structures 11 . Dissection along the avascular plane between the gland and the adjacent tissue with bipolar coagulation forceps or ultrasonic scalpels was easy because of the loose attachment between the surface of the gland and the connective tissue. The application of ultrasonic scalpels decreased the bleeding effectively, which provided a clear surgical field for operators. During endoscope-assisted operations the marginal mandibular branches of the facial nerve in all patients were well preserved because there was adequate distance between the dissection plane and the branch of nerve. Although sometimes the submandibular epidermoid cyst could be large the dissection along the surface of cyst was performed without difficulty. If the cyst was too big to go through the incision, suction of contents was performed to reduce the volume of the cyst. During the submandibular sialadenectomy the exposure of the loop of the lingual nerve, Wharton's duct, and the accompanying vessels below the  Table 2. Statistical analyses of all of the data. mylohyoid muscle plane was the key step. Every time we dissected this area the use of a 30° 4-mm endoscope and a 0° 4-mm endoscope were combined to secure a good surgical view as reported by other surgeons 4 . Some authors have reported that severe adhesion to the adjacent tissue existed in patients with submaxillaritis accompanying inflammation and as a result heavy bleeding took place during the procedure. Also, the lingual nerve was easily damaged in this situation due to the unclear working space 14 . Therefore, in patients with submaxillaritis an enhanced CT scan must be performed to exclude patients with inflammation 39 . How to resect space-occupying lesions in front of parotid gland or near the accessory parotid gland is controversial clinically. The use of the Blair incision for the removal of lesion in front of the parotid gland results in a long and obvious scar, although the procedure provides adequate exposure of the operative field 40 . Intraoral incisions might get a better cosmetic result compared with extraoral incisions 41,42 , however there are many drawbacks with intraoral incisions such as the poor exposure of operation field, surgical wound infections and inconvenience for eating 43 during recovery. In our study, a preauricular incision behind the ridge of the tragus was designed to extend downward along the crease between the ear and face to the interior margin of the earlobe. Complete resection of tumor with minimal disruption to the surrounding healthy tissue conformed to the principles of being minimally invasive 44 . This incision for EAA minimized the adverse effect of the scar visibility by using the natural structures around ear. For patients with an accessory tumor the smaller incision and longer distance from the lesion could increase the difficulty of the operation due to the challenge to realize the partial resection around the tumor 26 . However, resection of intermuscular vascular malformation in front of the parotid gland can be performed easily along the avascular plane around the pseudocapsule without fear of the incomplete removal of the pseudocapsule and satellitosis around the tumor.
There are still some limitations to the use of EAA. First, it is more time-consuming than EA due to the associated learning curve for perfecting the endoscope-assisted technique. Secondly, issues such as maintaining a clear a working space for the surgeon and how to adjust the endoscopes to the most suitable position need to be resolved. Thirdly, in cases of severe inflammatory adherences to the surrounding structures, the risk of vascular and neurological injury are increased.

Conclusion
In this study we performed operations with EAA and EA on patients with NSOL in oral and maxillofacial areas and achieved complete resection of the NSOL. Patients with EAA achieved minimal invasion and good cosmetic results compared with those with EA. Endoscope-assisted resection of NSOL is more practicable for the beginner and so is promising as a standard procedure in the oral and maxillofacial areas.