Abstract
Objectives A variety of endonasal sellar repair techniques have been described; many are complex, multilayered, and carry potential morbidity. We propose an effective, technically simple single-layer repair for select sellar defects, including those with an intraoperative cerebrospinal fluid (CSF) leak. Our technique utilizes only a synthetic dural substitute inlay and dural sealant glue without packing or lumbar drainage.
Design This is a retrospective review-based study.
Setting This study was conducted at tertiary care center.
Participants Patients who underwent endoscopic transsphenoidal surgery for pituitary adenoma and sellar reconstruction with the aforementioned inlay technique. Patients were selected for this technique if they had an identified intraoperative CSF leak, a patulous diaphragm (expanded and thinned diaphragma sella), or a comorbidity excluding them from a simpler onlay only reconstruction.
Outcome Measures Postoperative CSF leak and sinonasal morbidity included in the study.
Results A total of 409 subjects were identified; 368 were initial resections. Gross total resection of the pituitary adenoma was achieved in 356 (87.0%) cases. Average tumor size was 2.6 ± 1.1 cm. Average tumor volume was 10.8 ± 12.1 cm 3 . There were 135 intraoperative CSF leaks and 196 patulous diaphragms. There were five postoperative CSF leaks (1.2%), all of which occurred in the first half of our series. Pre- and postoperative sino-nasal outcomes test-22 scores were 19.2 ± 18.2 and 18.8 ± 21.3 ( p = 0.492), respectively.
Conclusion A synthetic dural substitute inlay and dural sealant glue is an excellent single-layer repair for sellar defects, even those with an intraoperative CSF leak. This technique is highly effective in preventing postoperative CSF leaks and does not utilize packing or lumbar drainage. It also avoids the potential cost and morbidity associated with more complex and multilayered closures.
Keywords: endoscopic, transsphenoidal, pituitary adenoma, sellar reconstruction, cerebrospinal fluid leak
Introduction
Over the last two decades, use of the endonasal endoscopic approach to the sella has increased substantially. In many centers around the United States, the endonasal endoscopic approach is the primary procedure used for resection of pituitary and other sellar tumors. 1 2 Postoperative cerebrospinal fluid (CSF) leakage, which occurs in an estimated 2 to 5% of patients, remains a major potential source of morbidity following endonasal pituitary surgery. 3 4 Therefore, a critical part of endoscopic transsphenoidal pituitary surgery is adequate sellar reconstruction.
A wide variety of techniques for sellar reconstruction have been described, all with the primary goal of reducing the rate of postoperative CSF leaks. While many reconstructive algorithms exist, few incorporate a simple and single layer closure for patients with an intraoperative CSF leak. 5 6 Furthermore, the currently described single-layer repair techniques often utilize additional adjuncts, including packing or lumbar drainage. 7 As a result, many patients with an intraoperative CSF leak are reconstructed with excessive techniques, potentially subjecting them to undue morbidity, operative time, cost, and risk of infection. 5 6 8
We present our institutional experience with the use of a single-layer inlay of synthetic dural substitute and dural sealant glue for the repair of sellar defects in select patients, particularly those with an identified intraoperative CSF leak. This novel technique does not require nasal packing or the use of a lumbar drain.
Methods
A retrospective chart review was performed to identify patients who underwent endoscopic transsphenoidal surgery for pituitary adenoma between 2005 and 2018, and were reconstructed with an inlay of synthetic dural substitute with dural sealant glue.
The surgical technique of this repair involves an inlay of a single sheet of synthetic dural substitute placed directly against the inner surface of the sellar dural opening. The graft is made slightly larger than the dural defect ( Fig. 1 ). A thin layer of dural sealant glue is then applied over the epidural surface of the repair. No additional autologous tissue, packing material, or lumbar drainage is utilized.
Fig. 1.
Illustration of synthetic dural substitute inlay. (A) Defect with dural leaflets reflected and diaphragma sella visualized, (B) placement of inlay synthetic dural substitute, (C) intraoperative appearance with inlay graft in place, and (D) intraoperative appearance with dural sealant glue in place overlying synthetic dural substitute inlay graft.
Patients with a low-flow intraoperative CSF leak identified (specifically, a leak through the diaphragm that was not communicating with a ventricle) and/or a patulous diaphragm (expanded and thinned diaphragma sella herniating into or through the sellar floor defect) were reconstructed using our described single layer technique. In addition, some patients without an intraoperative CSF leak or a patulous diaphragm may be repaired with this technique if they have a comorbidity that precludes them from our institutions typical sellar reconstruction for “low-risk” cases—a single layer onlay oxidized cellulose sheet. 9 This is determined on a case-by-case basis. These comorbidities may include conditions associated with poor wound healing (i.e., Cushing's, poorly controlled diabetes mellitus), the need for postoperative radiation, pseudotumor cerebri, morbid obesity, and obstructive sleep apnea (OSA) requiring an early return to use of continuous positive airway pressure (CPAP). In these cases, the described repair is utilized even in the absence of a CSF leak or patulous diaphragm. Of note, some patients with a defect at the superior anterior aspect of the diaphragma sella at its junction with the tuberculum sella dura often required the addition of an onlay reinforcement due to the inability to achieve a water-tight repair with an inlay graft alone. At our institution, this onlay layer is most commonly in the form of a nephrogenic systemic fibrosis (NSF). These patients were excluded from this study and analyzed separately.
Patient demographic information and medical history, pre- and postoperative sinonasal evaluation, tumor characteristics, intraoperative findings, and specifics of the postoperative course were collected. Surgical details included the extent of tumor resection (gross total vs. subtotal), identification of an intraoperative CSF leak and, if present, whether it was a high-flow or low-flow leak, and the presence of a patulous diaphragm. In addition, postoperative complications including epistaxis-requiring intervention (packing or cautery), sinusitis-requiring treatment with antibiotics, synechiae formation, septal perforation, crusting, pulmonary embolism, meningitis, and most importantly, development of a postoperative CSF leak were recorded.
Results were analyzed using IBM SPSS Statistics for Mac, Version 25.0 (IBM Corp., Armonk, New York, United States). Descriptive statistical analysis was performed. Values were compared using paired t -tests when applicable. Statistical significance was set at p <0.05.
Results
A total of 409 patients were identified that met inclusion criteria. Of these, 216 (52.8%) were males and 193 (47.2%) were females. The median age was 53.1 ± 14.0 years. The average body mass index was 30.8 ± 6.8 kg/m 2. Additional medical comorbidities are included in Table 1 .
Table 1. Patient demographics, tumor characteristics, intraoperative details, and medical comorbidities.
| Number of patients (%) | |
|---|---|
| Male | 216 (52.8) |
| Average age | 53.1 ± 14.0 |
| Average BMI | 30.8 ± 6.8 |
| Smoker | 153 (37.4) |
| Tumor designation | |
| Microadenoma | 19 |
| Macroadenoma | 390 |
| Functioning | 71 (17.4) |
| Average tumor greatest dimension (cm) | 2.6 ± 1.1 |
| Average tumor volume (cm 3 ) | 10.8 ± 12.1 |
| Tumor type | |
| Primary | 368 |
| Recurrent | 41 |
| Extent of resection | |
| Partial | 53 |
| Gross total | 356 |
| Cavernous sinus involvement | 143 (35.0) |
| Optic nerve compression/abutment | 119 (29.1) |
| Intraoperative CSF leak | 135 (33.0) |
| Patulous diaphragm | 196 (47.9) |
| Medical comorbidities | |
| Hypertension | 174 (42.5) |
| Smoking | 153 (37.4) |
| Myocardial infarction | 6 (1.4) |
| Congestive heart failure | 3 (0.7) |
| Atrial fibrillation | 8 (2.0) |
| Deep vein thrombosis/pulmonary embolism | 17 (1.7) |
| Chronic obstructive pulmonary disease | 8 (2.0) |
| No identified comorbidities | 223 (54.5) |
Abbreviations: BMI, body mass index; CSF, cerebrospinal fluid.
n = 409.
In regard to tumor characteristics, 390 (95.4%) had a macroadenoma and 19 (4.6%) had a microadenoma. The average tumor size was 2.6 ± 1.1 cm in its greatest dimension, and average tumor volume was 10.8 ± 12.1 cm 3 as measured on preoperative imaging. Functional tumors were identified in 71 (17.4%) of the patients. Of the total number of patients, 368 (90.0%) had primary tumors, while 41 (10.0%) had recurrent tumors. Gross total resection was achieved in 356 cases (87.0%). There was documented cavernous sinus involvement in 143 patients (35.0%) and optic nerve compression/abutment in 119 patients (29.1%).
Of the 409 patients included in this review, 135 patients (33.0%) had an identified intraoperative CSF leak. All of the leaks were low-flow leaks through the thinned diaphragma sella; 10 of these were at the superior anterior aspect of the diaphragma sella at its junction with the tuberculum sella dura (2.4%). One hundred ninety-six patients (47.9%) had a patulous diaphragm without CSF leakage. Seventy-eight patients repaired with this technique did not have an intraoperative CSF or a patulous diaphragm. These patients were repaired with this technique for a variety of criteria including OSA requiring an early return to CPAP, concern for poor wound healing (i.e., Cushing's disease, poorly controlled diabetes mellitus, and prior radiation), anticipated radiation treatment, tumor invasion of the dura requiring a wide sellar dura resection, and recurrent surgery.
In regard to sinonasal outcomes, there were five postoperative CSF leaks identified (1.2%), all of which occurred in the earlier half of the series. We have not experienced a postoperative CSF leak with this approach in the past 5 years and over 200 consecutive cases. One postoperative leak was managed with a lumbar drain for 4 days with resolution of leak. The other four leaks were repaired endoscopically, two with the addition of a NSF nephrogenic systemic fibrosis onlay and two with abdominal fat packing, one with the addition of abdominal fascia inlay/onlay. The repairs with abdominal fat packing and the lumbar drain placement were all prior to 2010 and during the early development of our repair techniques. Preoperative sino-nasal outcomes test (SNOT)-22 score was 19.2 ± 18.2, and postoperative SNOT-22 score was 18.8 ± 21.3 ( p = 0.492). The average length of stay was 2.5 ± 2.0 days. Thirty-eight patients (9.3%) required readmission within 30 days of discharge, most commonly for metabolic and hormonal imbalances ( Table 2 ). Additional postoperative complications are shown in Table 3 . Average length of follow-up was 37 months (range = 1–140 months).
Table 2. Reasons for 30-day readmissions.
| Reason for readmission | Number of patients |
|---|---|
| Pleural effusion | 1 |
| Lower extremity DVT | 4 |
| Pulmonary embolus | 1 |
| Hyponatremia/SIADH | 7 |
| Hypernatremia/diabetes insipidus | 4 |
| Hypocortisolemia | 2 |
| Epistaxis | 5 a |
| Concern for vision change | 3 b |
| Migraine/uncontrolled pain | 3 |
| Dyspnea | 1 |
| Viral/bacterial illness (unrelated) | 3 |
| Nausea/vomiting | 1 |
| Debridement | 1 |
| Jaundice (unrelated) | 1 |
| Unknown | 2 |
Abbreviations: DVT, deep vein thrombosis; SIADH, syndrome of inappropriate antidiuretic hormone secretion.
n = 38.
Note: One patient has both SIADH and hypocortisolemia.
Two patients with epistaxis were controlled in the operating room, one was cauterized at bedside and observed, one resolved spontaneously without intervention, and one were managed with unspecified means.
One patient with vision change returned to the operating room for optic nerve exploration and decompression.
Table 3. Postoperative complications.
| Complication | Number of patients (%) |
|---|---|
| Synechiae | 26 (6.4) |
| Crusting | 30 (7.3) |
| Epistaxis | 5 (1.2) |
| Septal perforation | 8 (2.0) |
| Pulmonary embolism | 3 (0.7) |
| CSF leak | 5 (1.2) |
| Meningitis | 0 (0.0) |
| None | 281 (68.7) |
| Other | 8 (2.0) |
| Preoperative SNOT | 19.2 ± 18.2 |
| Postoperative SNOT | 18.8 ± 21.3 |
Abbreviations: CSF, cerebrospinal fluid; SNOT, sino-nasal outcomes test.
n = 409.
There were no instances of chronic crusting, ulceration within the sphenoid, infection, or nonhealing defects. In the majority of patients, remucosalization was noted on postoperative rigid endoscopic exam by 3 weeks after surgery ( Fig. 2 ).
Fig. 2.
Rigid nasal endoscopy 1 week postoperatively demonstrating sellar healing and remucosalization.
Discussion
As a growing number of centers embrace endonasal endoscopic techniques for pituitary surgery, there is notable variability in reconstructive approaches, both with respect to technique and philosophy. The majority of described sellar reconstructive techniques are multilayered. While many are effective in preventing postoperative CSF leaks, they are often overly complex, time-consuming, expensive, and may result in undue morbidity. 5 6 8 These overly complex techniques are routinely incorporated into algorithms, such that a major limitation of existing graded reconstructive algorithms is that “low- and moderate-risk” patients receive robust reconstructions that are not necessarily warranted. Even existing descriptions of single layer reconstructions in the literature for patients with intraoperative CSF leaks seem overly complex as they utilize the addition of packing, lumbar drainage, or autologous mucosal tissue grafting. 7 In this study, we show that patients with an intraoperative CSF leak can be safely and effectively repaired using only a simple and single layer reconstruction with an inlay graft of synthetic dural substitute and dural sealant glue without the addition of packing or lumbar drainage.
In a previous publication, we have shown that “low-risk” patients, defined within our algorithm as those without an identified intraoperative leak, without a patulous diaphragm, and without prohibiting medical comorbidities are adequately repaired with just an onlay of layered sheets of regenerated oxidized cellulose. 9 We approach all of our patients, even those with an intraoperative CSF leak or a patulous diaphragm with similar simplicity of favoring a single layer reconstruction. Our institutional experience suggests that there is a large subset of patients undergoing pituitary surgery with these intraoperative findings for whom a simple and single layer inlay repair using an inlay sheet of synthetic dural substitute and dural sealant glue is very effective at preventing postoperative CSF leaks and carries low morbidity.
Within our institutional reconstructive algorithm, patients are primarily considered for this simple inlay sellar repair if they have an identified low-flow intraoperative CSF leak, a patulous diaphragm, or a comorbidity excluding them from the “low-risk” designation as described above. These comorbidities include anticipated poor wound healing (i.e., poorly controlled diabetes mellitus or Cushing's disease), severe OSA requiring an early resumption of CPAP, revision surgery, extensive dural invasion and subsequent wide resection, a history of prior radiation, or anticipated postoperative radiation. As such, the reconstructive technique is most often decided at the time of surgery based on the intraoperative findings. Of note, certain low-flow CSF leaks, specifically those that are at the anterior superior aspect of the diaphragma sella at its junction with the tuberculum sella dura are not always amenable to our simple inlay repair. This is due to the low probability of obtaining a water-tight seal with an inlay alone, as this opening is above the attachment of the diaphragma sella and directly into the suprasellar cistern through the dura. In our experience, these particular defects frequently require the addition of an onlay graft, most commonly a NSF onlay at our institution. When this scenario is encountered, the repair technique is converted intraoperatively from a single-layer reconstruction to a multilayered reconstruction as needed. Additionally, some comorbidities are considered so significant that patients are deemed “high-risk,” and are not considered appropriate candidates for this single layer inlay reconstruction. These are patients considered to be at particularly high risk of postoperative CSF leakage and often involve revision cases with large dural defects. These determinations are made on a case by case. Despite these exclusions, 59.6% of all patients undergoing endoscopic transsphenoidal surgery for pituitary adenoma in our cohort were eligible for the aforementioned “simple” inlay reconstruction. An additional 27% of patients were “low-risk” and repaired with the simple onlay reconstruction of sheets of oxidized cellulose alone, as described above, such that fewer than 15% of our pituitary adenoma patients require a multilayered reconstruction with a NSF.
The simplicity of our single layer reconstruction spares many patients the morbidity, cost, and operative time associated with more complex, multilayered closures. This often includes the nasal septal donor site morbidity and olfactory dysfunction related to the harvest and inset of the NSF. 10 11 Of note, although Ye et al describe a similar use of synthetic dural substitute in patients with diaphragm violation from an unidentified source, their described technique also involves filling the sellar cavity with hemostatic gauze or gelatin sponge and again using gelatin sponge on the outside of the sellar region, and then reinforcing this multilayer construct with fibrin glue. Due to the additional layers of reconstruction, it is felt that their description of a “similar” technique is significantly different and much more extensive, involving intradural and extradural packing. Ye et al report a postoperative CSF leak rate of 1.9%. 7
The postoperative leak rate in our cohort of patients was 1.2% and compares favorably to that of several large endoscopic pituitary adenoma series, even those that routinely use free mucosal and pedicled vascularized flaps, some of which are as high as 5%. 8 12 13 14 This demonstrates that our repair technique is very effective when used appropriately.
It has been our experience that the sellar face heals remarkably well following reconstruction with this technique and demonstrates rapid remucosalization. Objectively, preoperative and 3-month postoperative SNOT-22 scores were not statistically different.
Our study includes a large cohort of patients with a substantial length of follow-up. However, there are some limitations of this study, including those inherent to a retrospective review of one institution and a single team. While selection bias is inherent to all retrospective reviews, we demonstrate that in appropriately selected patients the technique described herein is very effective. Additionally, long-term issues in certain patients may not have been captured, although are unlikely given our significant duration of follow-up (on average over 3 years). A larger prospective trial comparing this reconstruction with other techniques appropriate for moderate-risk defects could help determine the optimal repair technique with respect to CSF leak rate, cost, and complication profile.
Conclusion
The use of a synthetic dural substitute inlay and dural sealant glue is an excellent repair for sellar defects, including those with an identified intraoperative CSF leak through the diaphragm sella. This novel technique is highly effective in preventing postoperative CSF leaks and does not utilize packing or lumbar drainage. Furthermore, it avoids the potential higher relative cost and morbidity associated with more complex, multilayered closures.
Footnotes
Conflict of Interest None declared.
References
- 1.Jho H D, Carrau R L, Ko Y, Daly M A.Endoscopic pituitary surgery: an early experience Surg Neurol 19974703213–222., discussion 222–223 [DOI] [PubMed] [Google Scholar]
- 2.Svider P F, Keeley B R, Husain Q. Regional disparities and practice patterns in surgical approaches to pituitary tumors in the United States. Int Forum Allergy Rhinol. 2013;3(12):1007–1012. doi: 10.1002/alr.21216. [DOI] [PubMed] [Google Scholar]
- 3.Tabaee A, Anand V K, Barrón Y. Endoscopic pituitary surgery: a systematic review and meta-analysis. J Neurosurg. 2009;111(03):545–554. doi: 10.3171/2007.12.17635. [DOI] [PubMed] [Google Scholar]
- 4.Cheng Y, Xue F, Wang T Y. Analyses and treatments of postoperative nasal complications after endonasal transsphenoidal resection of pituitary neoplasms. Medicine (Baltimore). 2017;96(15):e6614. doi: 10.1097/MD.0000000000006614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Jalessi M, Sharifi G, Mirfallah Layalestani M R. Sellar reconstruction algorithm in endoscopic transsphenoidal pituitary surgery: experience with 240 cases. Med J Islam Repub Iran. 2013;27(04):186–194. [PMC free article] [PubMed] [Google Scholar]
- 6.Kuan E C, Yoo F, Patel P B, Su B M, Bergsneider M, Wang M B. An algorithm for sellar reconstruction following the endoscopic endonasal approach: A review of 300 consecutive cases. J Neurol Surg B Skull Base. 2018;79(02):177–183. doi: 10.1055/s-0037-1606293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ye Y, Wang F, Zhou T, Luo Y. Low complication rate of sellar reconstruction by artificial dura mater during endoscopic endonasal transsphenoidal surgery. Medicine (Baltimore) 2017;96(52):e9422. doi: 10.1097/MD.0000000000009422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Conger A, Zhao F, Wang X. Evolution of the graded repair of CSF leaks and skull base defects in endonasal endoscopic tumor surgery: trends in repair failure and meningitis rates in 509 patients. J Neurosurg. 2018;130(03):861–875. doi: 10.3171/2017.11.JNS172141. [DOI] [PubMed] [Google Scholar]
- 9.Chaskes M B, Fastenberg J H, Vimawala S. A simple onlay sellar reconstruction does not increase the risk of post-operative cerebrospinal fluid leak in well-selected patients. J Neurol Surg B Skull Base. 2021;82 03:e231–e235. doi: 10.1055/s-0039-3402000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Soudry E, Psaltis A J, Lee K H, Vaezafshar R, Nayak J V, Hwang P H. Complications associated with the pedicled nasoseptal flap for skull base reconstruction. Laryngoscope. 2015;125(01):80–85. doi: 10.1002/lary.24863. [DOI] [PubMed] [Google Scholar]
- 11.Soyka M B, Serra C, Regli L, Meier E, Holzmann D. Long-term olfactory outcome after nasoseptal flap reconstructions in midline skull base surgery. Am J Rhinol Allergy. 2017;31(05):334–337. doi: 10.2500/ajra.2017.31.4463. [DOI] [PubMed] [Google Scholar]
- 12.Fraser S, Gardner P A, Koutourousiou M. Risk factors associated with postoperative cerebrospinal fluid leak after endoscopic endonasal skull base surgery. J Neurosurg. 2018;128(04):1066–1071. doi: 10.3171/2016.12.JNS1694. [DOI] [PubMed] [Google Scholar]
- 13.Pereira E AC, Grandidge C A, Nowak V A, Cudlip S A. Cerebrospinal fluid leaks after transsphenoidal surgery - Effect of a polyethylene glycol hydrogel dural sealant. J Clin Neurosci. 2017;44:6–10. doi: 10.1016/j.jocn.2017.06.016. [DOI] [PubMed] [Google Scholar]
- 14.Halvorsen H, Ramm-Pettersen J, Josefsen R. Surgical complications after transsphenoidal microscopic and endoscopic surgery for pituitary adenoma: a consecutive series of 506 procedures. Acta Neurochir (Wien) 2014;156(03):441–449. doi: 10.1007/s00701-013-1959-7. [DOI] [PubMed] [Google Scholar]