Rigid Autologous Bone Reconstruction after Endoscopic Endonasal Olfactory Neuroblastoma Resection

Background : For endoscopic endonasal reconstruction of skull base defects, minimizing the sequelae of Cerebrospinal Fluid (CSF) leakage, graft migration, and infection is challenging. This study aims to evaluate the outcomes of endoscopic multilayer autologous rigid buttress reconstruction in patients with olfactory neuroblastoma. Methods : This study retrospectively reviews the records of olfactory neuroblastoma patients who underwent endoscopic endonasal surgery and reconstruction with a rigid buttress at one institution between May 2015 and December 2021. After tumor resection, reconstruction was performed using a multilayered fascia lata and autologous rigid buttress, and the defect was covered with a vascularized nasoseptal flap. Autologous bone chips or septal cartilage was placed between the dura and bone of the skull base. We reviewed demographic data, postoperative complications, body mass index, the follow-up time, and tumor Kadish stage and Hyams grade. Results : Reconstruction was performed in 12 patients (4 males; 8 females). The median follow-up period was 24.8 months. 3 patients with Kadish stage D tumors with lymph node invasion underwent neck dissection surgery. Hyams tumor grade classification showed 5 patients with tumor grade II, 6 with grade III, and 1 with grade IV. Lumbar drainage was placed in one patient for monitoring (patient #1). 4 patients underwent radiotherapy, and 4 underwent chemoradiotherapy. No postoperative CSF leakage occurred in any patient. Conclusions : Reconstruction of skull base defects using a multilayered anatomical approach with an autologous rigid buttress and vascularized flap is effective and can prevent complications in patients after endoscopic endonasal surgery for olfactory neuroblastoma.


Introduction
Endoscopic approaches are now widely used in the management of skull-base neoplasms, including pituitary tumors, Olfactory Neuroblastomas (ONB), chordomas, and chondrosarcomas. Surgical treatment planning for such neoplasms requires consideration of the tumor extension and the risks of subsequent skull-base defects, postoperative Cerebrospinal Fluid (CSF) leakage, meningitis, brain abscess, and pneumocephalus [1,2] Because insufficient repair is associated with postoperative complications, reliable skull-base reconstruction after endoscopic surgery is necessary [3,4]. Reconstruction of anterior skull base defects with additional rigid structural support is essential to prevent graft migration due to intracranial pulsation and gravity and to reduce the possibility of frontal lobe herniation [5][6][7]. To the best of our knowledge, no previous study has investigated postoperative complications following endonasal endoscopic reconstruction with an autologous rigid buttress in ONB patients. This study aims to evaluate the results of endoscopic reconstruction using an anatomical multilayered approach with a rigid buttress and vascularized flap in ONB patients.

Patient Data
In this retrospective cohort study, we reviewed the records of ONB patients who underwent endoscopic endonasal surgery with endoscopic rigid reconstruction between May 2015 and December 2021. The diagnosis of ONB was histologically confirmed in all cases. We reviewed patient data including age, sex, diagnosis, surgical procedure, repair technique, length of follow-up, and the presence of postoperative complications. The final cohort included 12 patients. This study was approved by the Institutional Ethics Committee of Chang Gung Memorial Hospital. All cases were jointly managed by the same surgical team, which consisted of otorhinolaryngology and neurosurgery services and an experienced primary surgeon. The otorhinolaryngologist resected tumors located in the nasal cavity and paranasal sinuses, and the neurosurgeon resected dural, olfactory bulb, and intracranial tumors. All patients were evaluated preoperatively using enhanced computed tomography and/or magnetic resonance imaging and positron emission tomography ( Figure 1). Biopsied tumor samples were analyzed by a pathologist. The arrows indicate the fat graft. The arrowheads indicate the bone graft.

Surgical Procedures and Reconstruction Techniques
We identified and completely resected the tumor attachment. The cribriform plate, dura, olfactory bulbs, and tracts were resected depending on the extent of the tumor via navigation assistance. The skull base was then reconstructed. We harvested approximately 2.5 × 2.5 cm of the fascia lata, depending on the defect size, and subcutaneous fat from the thigh before reconstruction. An autologous fat graft was used to eliminate intracranial dead space and prevent the fascia graft from being dislodged into the space. For the first layer, a piece of autologous fascia lata was placed as the subdural inlay, and the margin was extended circumferentially about 0.5 cm beyond the edge of the dural defect. For the second layer (extradural onlay), another piece of autologous fascia lata larger than the bony defect circumferentially was placed outside the dural defect but inside the bony edge (in the epidural space). For the third layer

Postoperative Management
Lumbar drainage was initially inserted in one patient (patient 1) to prevent postoperative CSF leak (for monitoring only). We then considered that this additional procedure to prevent CSF leakage was not necessary. Therefore, midway through the study period we revised our protocol to exclude the lumbar drainage, which facilitated earlier mobilization of the patients. The patients were kept on a prophylactic antibiotic regimen consisting of a third-generation cephalosporin or a penicillin-based antibiotic (a beta-lactamase) from 1 day before until 7 days after the surgery.

Results
The clinical characteristics of the patients are shown in Table  1. Of the 12 included patients, 4 (33.3%) were male and 8 (66.6%) were female. Eleven patients had complete reconstruction with a bone graft, and 1 patient had incomplete reconstruction due to inadequate bone reconstruction. The mean patient age was 50.9 years (range, 27-78 years), and the mean postoperative followup period was 24.8 months (range, 7-57 months). The mean Body Mass Index (BMI) was 22.3 kg/m 2 . Three patients with Kadish stage D tumors and lymph node invasion underwent neck dissection surgery. Hyams tumor grading revealed a grade II tumor in 5 patients, grade III tumor in 6 patients, and grade IV tumor in 1 patient. The mean hospital stay for the 9 patients without neck dissection was 8.8 days, compared to 17.3 days for the 3 patients with neck dissection. All patients underwent postoperative endoscopic examination upon return to the otolaryngology clinic. No cases exhibited persistent crusting at the repair site, and there were no cases of postoperative meningitis, mucocele formation, or tension pneumocephalus. Four patients underwent radiation therapy, and 4 patients underwent chemoradiation therapy. One patient (patient #9) had vomer bone erosion leading to inadequate rigid reconstruction, which was managed with complete bed rest for 5 days. The patient did not undergo a second reconstructive surgery. Endoscopic examination was performed by an otolaryngologist every month after surgery in patients treated using the Valsalva maneuver to check for CSF leakage. In a total of 26 months of follow up, no leakage was noted among these patients. In the total patient cohort, we observed no CSF leakage, encephalocele, or central nervous system infection. The mean defect width was 1.56 cm, and the mean defect length was 1.97 cm. The cancer recurred in 3 patients during the follow-up period, as follows: Patient #2, inferior right nasal septum metastasis 23 months after tumor resection; patient #3, lumbar spine bone metastasis 55 months after radiotherapy; patient #4, right lower section of the inferior turbinate metastasis 26 months after concurrent chemoradiotherapy.

Discussion
Our study had described the experience and results of endoscopic reconstruction using an anatomical multilayered approach with a rigid buttress and vascularized flap in ONB patients. ONBs are rare, anatomically complex neuroendocrine malignancies that are thought to arise from the nasal olfactory epithelium. Due to this site of origin, they have a tendency to extend directly into the central nervous system through the cribriform plate and anterior skull base. No uniform treatment algorithm has been established for patients with ONB; however, multidisciplinary management including surgery alone or surgery with radiotherapy is generally preferred [8][9][10][11]. While surgical resection can be the sole therapy for early tumors, combination treatment is the preferred method for later tumors at most institutions [12]. The 5-year overall survival rate is 86-100% in patients with Kadish stage A and B tumors and 70.8-76% in patients with Kadish stage C and D tumors [13][14][15].

The Era of Endoscopic Surgery for Patients with ONB
En-bloc resection through a craniofacial approach has been the traditional technique used in the management of ONB. However, this procedure results in a large facial wound and requires a long hospitalization and intense postoperative care [2,16,17]. The endonasal endoscopic approach avoids the facial incision, extensive osteotomies, and frontal lobe retraction and is therefore less invasive with better preservation of function. The endonasal endoscopic approach is also reported to achieve better survival and local control rates [16,18,19]. A meta-analysis of 36 studies with 609 ONB patients suggests higher disease-specific survival and lower rates of CSF leak and intracranial complications in patients undergoing endoscopy compared to those undergoing craniofacial surgery [16]. The origin of the olfactory neuroblastoma is the olfactory neuroepithelium located in superior nasal vault and cribriform plate. Given this presumed origin, the appropriate surgical margin is considered to be resection of the dura and the olfactory apparatus. Our present retrospective study reveals no association between the recurrence site and reconstructed tissue. We had also reviewed the literature regarding tissue reconstruction and tumor recurrence. According to Florent Carsuzaa et al., [20] the rate of local recurrence after surgery for malignant sinonasal tumors was about 30-50%. To date, few studies have addressed tumor recurrence in reconstructed tissue in ONB patients.

Common Complications Of Endonasal Endoscopic Surgery
The complications of endonasal endoscopic surgery, including nasal obstruction, CSF leak, meningitis, pneumocephalus, and intracranial hypotension, may be troublesome and are often related to inadequate reconstruction [6,21,22]. Surgical site mobility may require further CSF diversion techniques or revision surgery, resulting in delayed treatment. Postoperative CSF leakage has been reported in 1-8% of anterior skull base reconstructions [5,23,24].

Differences in Endonasal Endoscopic Reconstruction Techniques
Multilayer reconstruction to establish watertight closure is the gold standard for skull base defects [4,25,26

The Importance Of A Rigid Buttress In Reconstruction
Radical tumor excision has been shown to be beneficial for overall survival. [35,37] However, oncological resection may result in large bony defects that require rigid reconstruction. The use of bony grafts to repair large defects in high-pressure areas or in cases of intracranial hypertension provides resistance to pulsation and pressure while allowing for healing and mucosalization. [38,39] Some studies recommend rigid reconstruction in patients at higher risk of postoperative CSF leak and meningocele. Such risk factors include elevated BMI, higher CSF pressure, larger bony and arachnoid defects, previous radiation, and future adjuvant therapy. From our point of view, for patients requiring further adjuvant treatment, better mucosal lining and healing are also considerations for rigid repair.
[25] A rigid buttress is more effective and facilitates mucosal flap ingrowth and epithelization, which also contribute to wound closure after reconstruction. Moreover, a rigid buttress also avoids the need for a Foley balloon, which has inherent risks of overinflation, local infection, and postoperative distress to the patient. [28,40,41] Whether bony reconstruction is necessary in all patients undergoing anterior skull base reconstruction is debatable.
[36] A rigid buttress is believed to prevent dislodgement. [42,43] Furthermore, spontaneous bony regrowth and osseous closure have been reported in rats, rabbits, and dogs.
[30] In addition, El-Banhawy et al. reported that after repair with free bone grafts in 13 patients, the grafts were incorporated into the bony skull base in 84.6% of the patients.
[44] These studies highlight the importance of rigid reconstruction and suggest that it is preferable to nonrigid reconstruction to avoid complications and graft migration.

Radiation As A Risk Factor for Reconstruction Failure
Comprehensive therapy is recommended for patients with ONB, including radiation for high-risk patients. In general, radiation therapy is recommended for postoperative treatment in patients who undergo non-radical surgery for T3 and T4, Kadish C and D, and high Hyams grade tumors.
[45] The long-term effect of radiation on autologous flaps in ONB patients is unclear. The detrimental effects of radiation therapy are reported to be caused by vascular damage and endothelial cell injury. Postoperative radiation toxicity also causes the tissue to become fibrotic, leading to radiation-induced chronic nonhealing wounds. Accordingly, Herle et al. suggest that the defect should be reconstructed using a rigid technique before radiotherapy.
[46] From our point of view, a rigid buttress plays a vital role in reconstruction for ONB patients, since they are likely to undergo radiotherapy.

Reconstruction Materials
The desired features of reconstruction materials are strength, the ability to revascularize, and biotechnical compatibility. Alloplastic implants such as titanium, porous polyethylene (MEDPOR), and silicone are foreign buttresses that may be susceptible to implantrelated infection, hemorrhage, distortion of magnetic resonance images, and the formation of granulation tissue.
[43] Bioresorbable alloplastic implants have adequate strength and flexibility to fit the defect while having a lower potential risk compared to permanent alloplastic buttresses, therefore offering an alternative choice for reconstruction. Disadvantages including longer resorption time, high cost, and image disruption have been reported. [28,40,38] Seaman et al. reported on 24 patients with extrasellar skull base tumors who underwent reconstruction with bioabsorbable plates.
[40] In 19 patients who received postoperative lumbar drainage, the postoperative CSF leak rate was 0%; however, this study did not report on the need for further radiation treatment or long-term infection rates.
[40] Autologous grafts remain a popular choice, and the use of harvested bone or septal cartilage for rigid buttresses in skull base reconstruction has several advantages, including fewer adverse immunologic reactions, greater biocompatibility, lower cost, lower risk of infection, and lower impact on adjuvant therapy and image interpretation. [43,44] In particular, the production of fibroblasts and angiogenic factors in the acellular dermis can be faster in autografts than to allografts, therefore resulting in a shorter period of healing and wound crusting.
[43] In addition, Jin et al. reported that bone in situ reconstruction can prevent adhesion of the nasoseptal flap to intracranial nerves or vessels, providing safe access for patients who require reoperation. [7] In this study, we introduce a technique for repairing skull base defects with an autologous rigid buttress and vascularized nasoseptal flap using a multilayered approach. This method achieved remarkable results in our ONB patients.

Limitations
This study has several limitations. First, selection bias is possible with regards to the selection of suitable candidates for endoscopic surgery. Thus, direct comparison between the results of our study and other reconstruction techniques is difficult. Second, given the relative rarity of the disease and approach, it is important to evaluate both intraoperative and postoperative complications. Long-term follow-up of surgical site mobility and the effect of post-radiation therapy should also be taken into consideration. Third, a larger sample size is needed to evaluate the oncological outcomes.

Conclusions
The use of a multilayered approach to autologous rigid graft reconstruction preserves anatomical structures. In addition, this technique avoids the intraoperative complications of endoscopic surgery, provides longstanding safety and adequate watertight closure, facilitates defect healing, and enables an earlier return to daily activities and adjuvant therapy.