Graded Reconstruction Strategy Using a Multi-Layer Technique without Lumbar Drainage After Endoscopic Endonasal Surgery

Sellar following surgery (EES) requires modication based on the degree of cerebrospinal uid (CSF) leak. For high-ow (grade II or III) intraoperative CSF leak, lumbar drainage (LD), in addition to the multi-layer closing technique, is generally recommended. However, LD has complications occasionally, including post-puncture headache, over-drainage symptoms, and increased length of stay (LOS). We retrospectively evaluated the outcome of our graded reconstruction strategy using a multi-layer technique with a novel material, without LD, after EES. Methods

Introduction especially in cases with a high body mass index (BMI), opening of the third ventricle, previous history of surgery or radiation treatment, or involvement of the posterior cranial fossa [14][15][16][17]. However, LD is frequently associated with unlethal complications, such as post-puncture headache, radiculopathy, symptoms associated with over-drainage, and immobilization [18], all of which could increase the postoperative length of stay (LOS). Therefore, the development of safe and effective alternative reconstruction methods that do not require LD is necessary to reduce the incidence of LD-associated complications.
Injectable hydroxyapatite (HXA) was introduced in 1996 and is has been widely used in various elds, including orthopedic surgery, otorhinolaryngology, plastic surgery, and neurosurgery [19][20][21]. The HXA hardens within 5 min after injection under drying conditions; hence, it could be a suitable material to cover bony defects. In skull base surgery for sellar reconstruction, Chung et al. recommended injectable HXA as a substitute for NSF [7], and Kitano et al. proposed injectable HXA as a component of the multilayer technique [22]. In this study, we retrospectively evaluated the outcome of our graded reconstruction strategy after EES with injectable HXA, but without LD.

Materials And Methods
A total of 250 patients who underwent EES between June 2020 and March 2021 were retrospectively reviewed in this consecutive series. During this period, we performed EES for skull base tumors such as pituitary adenoma, craniopharyngioma, tuberculum sellae meningioma, and Rathke's cleft cyst. The degree of intraoperative CSF leak was assessed using the grading scale suggested by Esposito et al [17].
Ninety-seven patients showed an intraoperative grade II or III CSF leak and were included in the study.
Patients without an intraoperative CSF leak, patients with a grade I CSF leak, and those with tumors involving the posterior fossa were excluded from this study. The study was approved by the institutional review board of our institution and was performed in compliance with relevant ethical guidelines.
Medical records of the included patients were reviewed to collect data, including demographic variables (sex, age, and BMI), diagnosis, intraoperative CSF leak grade, sellar reconstruction method used, and length of stay (LOS) in the hospital postoperatively. To investigate the integrity of NSF, postoperative MRI and rhinological status were also examined by a dedicated neuroradiologist and otorhinolaryngologist, respectively.

Surgical technique
Details of the EES have been described previously [8,23]. The entire surgery was performed by an endoscopic skull base surgery team comprising a neurosurgeon (D.S.K.) and an otorhinolaryngologist (S.D.H.) at a single institution. During the EES, rigid endoscopes (diameter, 4 mm; length, 18 cm; with 0º, 30º, or 45º angle of view lenses; KARL STORZ Endoscopy Korea Co., Ltd., Seoul, Korea) were used in combination with a robotic holding arm (POINT SETTER, Mitaka Kohki Co., Ltd., Tokyo, Japan). All endoscopic procedures were recorded using a high-resolution camera and video recording system (KARL STORZ Endoscopy Korea Co., Ltd., Seoul, Korea).
For grade II CSF leak, we placed a brin sealant patch (TachoSil®, Takeda Pharmaceutical Co., Ltd., Osaka, Japan), followed by an NSF. In contrast, for cases involving grade III CSF leaks, the sellar reconstructive method included the following multi-layer technique. In this technique, (i) a collagen matrix (DuraGen®, Integra LifeSciences, New Jersey, USA) was rst placed into the arachnoid defect. Thereafter, (ii) an acellular dermal graft (AlloDerm®, BioHorizons®, Alabama, USA or MegaDerm®, L&C BIO, Gyeonggi-do, Korea), tailored according to the size of the sellar bone defect, was overlaid as an on-lay dura graft. (iii) HXA (Hydroset®, Stryker Leibinger, Freiburg, Germany) was then injected for additional closure of the sellar defect; at this step, no intraoperative CSF leak was observed from the reconstructed graft site 5-10 min after HXA injection. Finally, (iv) the NSF, already harvested at the beginning of EES, was used to cover the layered materials; thus, HXA was not exposed to the nasal cavity. Compression was applied with (v) a 12-Fr balloon catheter or two pieces of Merocel® (Medtronic, Minneapolis, USA) to stabilize the graft. Postoperatively, LD was not routinely performed.

Postoperative management
Three antibiotics (third-generation cephalosporin, quinolone, and metronidazole) were intravenously administered for 3-5 days. Intravenous or oral steroid (hydrocortisone) replacement was performed preand postoperatively. To identify the degree of tumor resection and the integrity of NSF, sellar MRI was performed one day after surgery, and otorhinolaryngological examination was performed on days 2 and 5 postoperatively. All patients were followed up at an otorhinolaryngology and endocrinology outpatient clinic at 1 week after discharge and at a neurosurgery outpatient clinic 3 weeks after discharge.

Statistical analysis
Independent variables associated with CSF leaks were analyzed using univariate analysis. The postoperative LOS was compared between the CSF leak grade II and grade III groups using an independent sample t-test. P < 0.05 was considered statistically signi cant. Statistical evaluations were accomplished by SPSS® Statistics 27 (IBM Co., New York, USA)

Baseline characteristics
Among the 97 consecutive patients included in this study, 47 (48.5%) were men and 50 (51.5%) were women. The mean age at the time of surgery was 48.5 (range, 5-75) years, and the mean BMI was 25.7 (range, 13.9-37.1) kg/m 2 . Pituitary adenoma (n = 57) was the most common tumor (including four cases of Cushing's disease, three cases of acromegaly, and three cases of prolactinoma) followed by craniopharyngioma (n = 18), meningioma (n = 11), Rathke's cleft cyst (n = 4), germinoma (n = 4), chordoma (n = 1), cavernous malformation (n = 1), and pilocytic astrocytoma (n = 1) ( Table 1).  (Table 3). We applied the modi ed reconstructive methods according to the CSF leak grade. We placed a brin sealant patch (TachoSil®, Takeda Pharmaceutical Co., Ltd., Osaka, Japan), followed by an NSF in patients with grade II CSF leak; for patients with grade III CSF leak, we applied a multi-layer technique using a collagen matrix, an acellular dermal graft, HXA, and an NSF. During the follow-up period, no postoperative CSF leak was observed in any of the patients. On univariate analysis, we found that no signi cant variable affected the outcome of the reconstruction.

Status of integrity of the NSF
To evaluate the integrity of the NSF, postoperative MRI was performed within a day after surgery. A Cshaped con guration of the ap adjacent to the skull base covering the sellar defect in both coronal and sagittal views was reviewed by an authoritative neuroradiologist (S.T.K.). Additionally, rhinological examination was performed by a dedicated otorhinolaryngologist. MRI results showed that NSF was not enhanced in 12 of the 97 patients examined (12.4%, 7 with grade II and 5 with grade III CSF leaks), which was con rmed by rhinological examination.

Adverse effects associated with HXAgraft
During the overall follow-up period, we did not observe crust, seroma, or infection around the NSF or HXA in routine rhinological examinations. In addition, there was no allergic reaction or aseptic meningitis in the group that received the HXA graft.

Discussion
Recently, EES for skull base tumors has become a major surgical technique in modern skull base surgery. Despite its surgical and technical limitations compared with the transcranial approach, it provides a favorable surgical outcome [24][25][26]. However, postoperative CSF leak, a recognized complication associated with EES [27,28], remains a matter of concern, and reconstruction of the barriers between the arachnoid space and sinonasal tract continues to be a challenge [10]. For grade III CSF leaks, which is an inevitable result after the suprasellar or transclival approach, lumbar CSF diversion can contribute to improved reconstruction outcomes [17,18]. However, LD often poses risks of associated complications such as post-puncture headache, radiculopathy, symptoms associated with over-drainage, and longstanding immobilization.
The recommended maintenance period of LD varies among studies [17,18], with no consensus. In clinical practice, postoperative LD should be maintained for a minimum of 3-5 days, resulting in prolonged LOS in hospitals. Therefore, an alternative technique without LD is required to manage grade III CSF leaks. Herein, we analyzed the e cacy of HXA and its potential to replace LD in high-ow intraoperative CSF leaks by comparing the LOS between patients with grade II and grade III CSF leaks. We found no signi cant difference in the LOS between the two groups, and no serious complications associated with injectable HXA were identi ed.
In contrast to a previous study on the complications of HXA [29], our study showed acceptable clinical results with HXA. For successful grafting of HXA, optimal condition of the perigraft environment is crucial because HXA weakens in humid environments and can fracture easily. Some materials, such as HXA, reabsorb too slowly, which can cause in ammation. Therefore, it is critical to maintain minimal or zero infection rate after HXA grafting. We suggest that the acellular dermis for dural grafting might function as a barrier between the arachnoid space and the injected HXA, and it may prevent stray HXA particles from entering the intradural space. In addition, in the study, placement of a vascularized NSF completely covered the grafted HXA and prevented it from exposure to the outer side of the sinonasal tract. Thus, the combination of acellular dermal graft with HXA could have contributed to our favorable results; no patient showed HXA-related complications, such as aseptic meningitis or foreign body reaction.
Our study had several limitations. Since this study was designed as a retrospective consecutive series, we did not compare the LOS between patients treated with injectable HXA and those who required LD. Furthermore, data were collected from a single institution by a single surgery team, and thus, might not represent the patient-group in general. Although the size of the tumor and the extent of the sellar defect varied among individual patients, these were not considered in our study.

Conclusions
The graded surgical repair strategy using injectable HXA in combination with acellular dermal grafts provided effectiveness and safety in sellar reconstruction after EES without postoperative CSF leak.
Combined use of injectable HXA and acellular dermal grafts for high-ow CSF leak can limit LD requirement, without signi cant risks. By decreasing the need for LD, HXA may allow early ambulation and shorten the LOS after surgery.