Abstract
The success of the endoscopic endonasal approach (EEA) to surgically manage clival chordomas (CC) relies on robust repair methods to reduce complications, such as cerebrospinal fluid (CSF) leaks. Our study aims to evaluate the existing literature to assess reconstructive techniques utilised and post-operative CSF leak rates in this cohort. A systematic review and analysis was performed of all published data related to CC patients managed with an EEA. A total of 24 articles were included, representing 363 patients and 396 procedures. A variety of reconstruction methods were used with 95.9% of studies using an intracranial repair graft, 70.8% using a nasoseptal flap (NSF), 62.5% using glue/haemostat, 58.3% using nasal packs and 75.0% employing multi-layered reconstruction. Post-operative CSF leak rate was 10.1%. The leak rate was less in subgroups where a NSF was used (9.4%) although this was not statistically significant (p = 0.273). There were no differences in leak rates when glue/haemostat (p = 0.139) or nasal packs (p = 0.550) were used. Our review is the most up-to-date synthesis of the existing literature surrounding the EEA to CCs assessing reconstruction and post-operative CSF leaks. It demonstrates most authors employ a multi-layered reconstruction method. The lack of statistical significance observed for CSF leaks in subgroups is likely due to a variety of cofounding surgeon and patient factors. Higher quality prospective randomised multi-centric studies, with reporting of specific repair techniques will enable future systematic reviews to provide a more accurate consensus regarding optimal methods of reconstruction in this field.
Keywords: Clival chordoma, Cerebrospinal fluid leak, Endoscopic surgery, Rhinology, Skull base, Skull base reconstruction
Introduction
Chordomas are rare tumours derived from notochordal remnants with an incidence of 0.08 per 100,000 [1]. They are classified as low-grade malignancies, although their infiltrative nature means a high rate of recurrence, and a median survival of 7.7 years [2]. Clival chordomas (CC) account for up to 40% of chordomas [3], with anatomical location posing a significant barrier to adequate resection and post-operative outcomes. Medical therapies such Imatinib or Erlotinib have been trialled in the treatment of CCs but have failed to demonstrate adequate tumour control [4].
Tumour extension is highly variable, ranging from restricted infiltration of the clival bone, to intra-cranial extension extending to the posterior fossa. For this reason, gross tumour resection (GTR) is surgically challenging [5]. Recent advances in the endoscopic endonasal approach (EEA) have largely superseded the more traditional aggressive transcranial approaches, providing superior visualisation, limiting retraction of neurovascular structures, permitting ventral exposure to deep lesions [6], but crucially permitting the use of vascularised mucosal flaps [7]. This has been made possible by improved endoscopic instrumentation as well as improved reconstructive understanding over recent years.
Although there are a variety of descriptions of techniques and approaches for resection of CCs, the rarity of presentation (0.1% of all skull base malignancies) [8] and the paucity of data at our disposal by means of either retrospective or prospective studies have been limited. This has led to no clear consensus regarding the ideal reconstructive technique that minimises post-operative complications. The success of these procedures relies on robust repair methods to reduce such complications, such as cerebrospinal fluid (CSF) leaks, meningitis and pneumocephalus. The incidence of CSF leak post-operatively remains the most important measure in the success of any repair.
Our study aims to evaluate existing literature through a systematic review to assess the current reconstructive techniques utilised and post-operative CSF leak rates using an EEA for resection of CCs.
Methods
Study Design
A systematic review and analysis was performed of all published data related to CC patients managed with an EEA where articles reported a detailed account of their reconstructive techniques and post-operative CSF leak rates. This was undertaken in accordance with the PRISMA guidelines [9]. The protocol for this review was registered on the PROSPERO database in March 2021.
Search Strategy
Electronic searches of the following databases were systematically undertaken in March 2021; EMBASE (1974 – March 2021), MEDLINE (1946 – March 2021), and the Cochrane Library. Limits used in the search included articles only published in English and conference abstracts and duplicates were excluded. Databases were accessed through the University of Hospitals Birmingham NHS Trust library. The search was conducted using Medical Subject Heading (MeSH) terms (Table 1). The reference lists of any relevant articles and systematic reviews were manually searched to identify any additional articles. Relevant articles already known to the authors were also included if not identified in the search.
Table 1.
Search strategy
| # | MeSH search term | Database |
|---|---|---|
| 1 | ("clival chordoma*").ti,ab | Medline |
| 2 | ("clivus chordoma*").ti,ab | Medline |
| 3 | ("skull base chordoma").ti,ab | Medline |
| 4 | (1 OR 2 OR 3) | Medline |
| 5 | (endoscop*).ti,ab | Medline |
| 6 | (4 AND 5) | Medline |
| 7 | ("clival chordoma*").ti,ab | EMBASE |
| 8 | ("clivus chordoma*").ti,ab | EMBASE |
| 9 | ("skull base chordoma").ti,ab | EMBASE |
| 10 | (7 OR 8 OR 9) | EMBASE |
| 11 | (endoscop*).ti,ab | EMBASE |
| 12 | (10 AND 11) | EMBASE |
| 13 | 12 [Publication types Article OR Editorial OR Letter OR Note OR Report OR Review] [English language] [Languages English] | EMBASE |
Study Selection
Primary screening was performed by two authors independently (K.K.G and E.B) with any discrepancies resolved by the senior author (K.J). Articles were initially excluded if their titles or abstracts were clearly irrelevant. The full texts of remaining articles were sought for eligibility screening. We included all articles that met the inclusion and exclusion criteria. Inclusion criteria were all articles that reported data for patients with CCs managed surgically with an EEA and reported data for their reconstructive methods and post-operative CSF leaks. Exclusion criteria were alternate surgical approaches (for example transcranial or combined approaches where data and outcomes were not separated for EEA), other pathologies (for example clival xanthoma), articles not in English and articles not reporting details regarding reconstruction methodology or post-operative CSF leak rates. Meta-analyses/systematic reviews and case reports were excluded as well as articles reported duplicated data.
Data Extraction and Analysis
Data was extracted from included papers by all four authors independently to a standard spreadsheet. Following tabulation, the data was re-evaluated against the original data source to avoid any clerical errors related to data entry. Extracted data included study characteristics (year, design, number of patients and procedures, follow-up) and patient characteristics (age, gender, GTR). Primary outcome measures were (1) reconstructive techniques used in terms of repair used (for example fat/fascia/duragen), nasoseptal flap (NSF), glue/haemostat, nasal packing, lumbar drain (LD) utilisation, whether these repairs were multi-layered or not, and (2) post-operative CSF leak rate. Other post-operative complications were also recorded.
Basic descriptive statistics were used where applicable and percentage values were calculated in each individual study for gender, GTR, LD utilisation, post-operative CSF leak and other complications. An overall percentage was calculated for each of these parameters. Sub-group analysis was performed for reconstruction methods used. This was used to compare the post-operative CSF leak rate between studies that utilised a NSF and those that did not. This was repeated for other reconstructive methods – i.e. glue/haemostat and for nasal packs. Statistical comparison between groups was achieved using chi-squared tests (χ2). A p-value of < 0.05 was deemed statistically significant in all applicable analysis. All calculations were completed using SPSS Statistics version 27.0.1.0 (IBM, Armonk, NY) and MS Excel® version 16.31 (Microsoft, Redmond, WA).
Results
Search Results
The study selection process is illustrated in Fig. 1. The literature search identified 213 articles after which 61 underwent eligibility screening following exclusion of duplicates and articles based on their titles and abstracts. A total of 24 full-length articles met our inclusion criteria published between 2005 – 2020 [6, 10–32]. This included a total of 363 patients (396 procedures) where CCs were treated with an EEA and details were reported in terms of post-operative CSF leak rate and reconstructive methods.
Fig. 1.
Study selection process of included articles
Table 2 summarises the study design and baseline characteristics of all included studies and participants. The majority of included studies were retrospective case series (n = 22, 91.7%) [6, 10–16, 19–32], with two studies reporting prospective data (8.3%) [17, 18]. The number of participants ranged from 2 [19, 20] to 65 [6]. Mean age ranged from 31 [27] to 62 [31] with reported GTR rates ranging from 0% [30] to 100% [24] with an average GTR of 55.1%. Mean follow up ranged from 6 [19, 20, 27] to 68 [10] months.
Table 2.
Baseline characteristics of included studies and participants
| Study and year | Study design | Patients (n) | Procedures (n) | Mean age (years) | Male (%) | GTR (%) | Follow up (months) |
|---|---|---|---|---|---|---|---|
| Oishi, 2020 [10] | Retrospective | 17 | 17 | Not reported (NR) | 10 (58.8) | 5 (29.4) | 68 |
| Wang, 2020 [11] | Retrospective | 49 | 49 | NR | 28 (57.1) | 32 (65.3) | NR |
| Soloperto, 2019 [12] | Retrospective | 8 | 8 | 61 | 4 (50.0) | 2 (25.0) | 25 |
| Yousaf, 2019 [13] | Retrospective | 15 | 15 | 49 | NR | 4 (26.7) | 40 |
| Rahme, 2018 [14] | Retrospective | 17 | 23 | 48 | 10 (58.8) | 9 (52.9) | 63 |
| Zoli, 2018 [6] | Retrospective | 65 | 80 | 48 | 34 (52.3) | 47 (72.3) | 52 |
| Shimony, 2017 [15] | Retrospective | 15 | 17 | 53 | 10 (66.7%) | 4 (26.7) | 20 |
| Catapano, 2016 [16] | Retrospective | 7 | 7 | NR | NR | 2 (28.6) | 39 |
| Garzaro, 2015 [17] | Prospective | 9 | 9 | 57 | 9 (100.0) | 6 (66.7) | 9 |
| Chibbaro, 2014 [18] | Prospective | 54 | 58 | NR | 33 (61.1) | 35 (64.8) | 34 |
| Iacoangeli, 2014 [19] | Retrospective | 2 | 2 | NR | NR | NR | 6 |
| Ishii, 2014 [20] | Retrospective | 2 | 2 | NR | NR | NR | 6 |
| Shidoh, 2014 [21] | Retrospective | 9 | 9 | 56 | 4 (44.4) | 3 (33.3) | NR |
| Saito, 2012 [22] | Retrospective | 6 | 6 | 59 | 3 (50.0) | 3 (50.0) | 16 |
| Tan, 2012 [23] | Retrospective | 14 | 14 | NR | 8 (57.1) | 7 (50.0) | 41 |
| Taniguchi, 2011 [24] | Retrospective | 4 | 4 | 57 | 2 (50.0) | 4 (100.0) | 21 |
| Fraser, 2010 [25] | Retrospective | 8 | 8 | 50 | 5 (62.5) | 4 (50.0) | 9 |
| Holzmann, 2010 [26] | Retrospective | 13 | 13 | 46 | 7 (53.8) | 11 (84.6) | 18 |
| Arbolay, 2009 [27] | Retrospective | 2 | 2 | 31 | 1 (50.0) | 1 (50.0) | 6 |
| Stippler, 2009 [28] | Retrospective | 20 | 26 | 44 | 10 (50.0) | 9 (45.0) | 13 |
| Dehdashti, 2008 [29] | Retrospective | 12 | 12 | 49 | 8 (66.7) | 7 (58.3) | 16 |
| Hwang, 2007 [30] | Retrospective | 3 | 3 | 52 | 3 (100.0) | 0 (0.0) | 40 |
| Frank, 2006 [31] | Retrospective | 9 | 9 | 62 | 3 (33.3) | 3 (33.3) | 27 |
| Solares, 2005 [32] | Retrospective | 3 | 3 | 50 | 2 (66.7) | 2 (66.7) | 13 |
| Totals/averages | – | 363 | 396 | 51 | 194 (53.4) | 200 (55.1) | 26 |
Outcome Measures
From the included studies, the post-operative CSF leak rate was 10.1% (n = 40). There were a wide variety of reconstruction methods used with 95.8% studies (n = 23) using an intracranial repair graft [6, 10–14, 16–32], 70.8% (n = 17) using a NSF [6, 10–14, 16–18, 20–26, 28], 62.5% (n = 15) using a glue/haemostat [32], and 58.3% (n = 14) using a nasal pack [6, 10, 12, 14, 16, 21–26, 28–30]. There were 20 studies (80%) using a multi-layered reconstruction technique [6, 10–14, 16–26, 28, 29]. Nine studies reported the use of a LD in a total of 55 patients (13.9%) [13, 15, 16, 18, 23–26, 29] The results from all outcome measures are summarised in Table 3. The overall number of other post-operative complications was 63 (15.9%). These were a variety of complications, the most common being cranial neuropathies and diplopia (n = 19, 30.2%), and meningitis (n = 15, 23.8%). The breakdown of all other post-operative complications, excluding CSF leaks, is summarised in Table 4.
Table 3.
Results from outcome measures in terms of reconstruction, post-operative CSF leaks and other complications
| Reconstruction | ||||||||
|---|---|---|---|---|---|---|---|---|
| Study and year | Repair | NSF (Y/N) | Glue/Haemostat | Nasal pack | Multi-layered reconstruction (Y/N) | Lumbar drain (%) | CSF leak (%) | Other complications (%) |
| Oishi, 2020 [10] | Fascia | Y | Oxidised cellulose, fibrin glue | Foley catheter | Y | 0 (0.0) | 2 (11.8) | 2 (11.8) |
| Wang, 2020 [11] | Fascia | Y | Nil | Nil | Y | 0 (0.0) | 6 (12.2) | 5 (10.2) |
| Soloperto, 2019 [12] | Fascia | Y | Nil | Foley catheter | Y | 0 (0.0) | 0 (0.0) | 1 (12.5) |
| Yousaf, 2019 [13] | Artificial dura | Y | Glue | Nil | Y | 15 (100.0) | 2 (13.3) | 2 (13.3) |
| Rahme, 2018 [14] | Fat | Y | Sealant | Foley catheter | Y | 0 (0.0) | 6 (26.1) | 5 (21.7) |
| Zoli, 2018 [6] | Fat, fascia, bone/cartilage | Y | Gelofoam | Merocel | Y | 0 (0.0) | 2 (2.5) | 9 (11.3) |
| Shimony, 2017 [15] | Nil | N | Nil | Nil | N | 9 (52.9) | 0 (0.0) | 5 (29.4) |
| Catapano, 2016 [16] | Fat, fascia | Y | Fibrin glue | Anterior nasal pack | Y | 7 (100.0) | 1 (14.3) | 2 (28.6) |
| Garzaro, 2015 [17] | Fat, fascia | Y | Fibrin and gelatin glue | Nil | Y | 0 (0.0) | 2 (22.2) | 2 (22.2) |
| Chibbaro, 2014 [18] | Fat, fascia | Y | Nil | Nil | Y | 4 (6.9) | 4 (6.9) | 8 (13.8) |
| Iacoangeli, 2014 [19] | Fat, fascia | N | Bone dust, fibrin glue, sutures | Nil | Y | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Ishii, 2014 [20] | Fat, fascia | Y | Nil | Nil | Y | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Shidoh, 2014 [21] | Fascia | Y | Oxidised cellulose, fibrin glue | Foley catheter | Y | 0 (0.0) | 1 (11.1) | 4 (44.4) |
| Saito, 2012 [22] | Fat, fascia | Y | Oxidised cellulose, fibrin glue | Foley catheter | Y | 0 (0.0) | 0 (0.0) | 3 (50.0) |
| Tan, 2012 [23] | Fat, fascia | Y (n = 5) | Gelofoam paste | BIPP ribbon | Y | 2 (14.3) | 3 (21.4) | 3 (21.4) |
| Taniguchi, 2011 [24] | Fat, fascia | Y | Nil | Sinus balloon | Y | 1 (25.0) | 0 (0.0) | 1 (25.0) |
| Fraser, 2010 [25] | Fat, duraseal, gasket seal | Y | Floseal | Tefla nasal splints | Y | 2 (25.0) | 0 (0.0) | 1 (12.5) |
| Holzmann, 2010 [26] | Fascia | Y (n = 12) | Nil | Nasal pack | Y | 11 (84.6) | 1 (7.7) | 2 (15.4) |
| Arbolay, 2009 [27] | Fat | N | Nil | Nil | N | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Stippler, 2009 [28] | Duragen | Y | Surgicel | Gelatin foam, nasal pack | Y | 0 (0.0) | 5 (19.2) | 5 (19.2) |
| Dehdashti, 2008 [29] | Fat, fascia | N | Fibrin glue | Gelofoam balloon | Y | 4 (33.3) | 4 (33.3) | 3 (25.0) |
| Hwang, 2007 [30] | Fat | N | Fibrin glue | Anterior nasal pack | N | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Frank, 2006 [31] | Fat | N | Nil | Nil | N | 0 (0.0) | 1 (11.1) | 0 (0.0) |
| Solares, 2005 [32] | Fat | N | Collagen, gelofoam, thrombin | Nil | N | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Totals/averages | – | – | – | – | 55 (13.9) | 40 (10.1) | 63 (15.9) | |
| N of studies reconstruction method used in (%) | 23 (95.8) | 17 (70.8) | 15 (62.5) | 14 (58.3) | 18 (75.0) | 9 (37.5) | – | – |
Table 4.
All post-operative complications excluding CSF leaks from all included studies
| Complication | n (%) |
|---|---|
| Cranial nerve palsy and diplopia | 19 (30.2) |
| Meningitis | 15 (23.8) |
| Endocrinopathies (including diabetes insipidus) | 7 (11.1) |
| Major haemorrhage | 6 (9.5) |
| Paresis (quad/hemi) | 4 (6.3) |
| Hydrocephalus | 2 (3.2) |
| Pneumocephalus | 1 (1.6) |
| Brain infarction | 2 (3.2) |
| Haematoma | 2 (3.2) |
| Epistaxis | 1 (1.6) |
| Craniocervical fixation | 1 (1.6) |
|
Medical complications Hypokalaemia Pulmonary embolus Pneumonia |
3 (4.7) 1 1 1 |
| Total | 63 (100.0) |
Subgroup Analysis
Subgroup analyses were performed to compare post-operative CSF leak rates according to reconstruction methods. Analysis for patients with LD was omitted as seven studies using LDs did not stratify post-operative CSF leak rates according to whether a LD was used or not. Where studies reported data where CSF leak rates were stratified according to reconstruction methods, these data were separated accordingly into each subgroup [23, 26].
The CSF leak rate was less than the overall CSF leak rate (10.1%) in subgroups where a NSF was used (9.4%), where no glue/haemostat was used (7.4%) and where no pack was used (9.0%). In terms of comparison between groups, there was a lower rate of CSF leak when a NSF was used compared to when one was not used (9.4% vs. 13.8%), but this was not a statistically significant difference (p = 0.273). There were higher rates of CSF leaks seen when glue/haemostat was used compared to not used (12.0% vs. 7.4%) and when nasal packs were used compared to not used (10.9% vs. 9.0%). Both differences were not statistically significant (p = 0.139 and 0.550 respectively). These findings are summarised in Table 5.
Table 5.
Subgroup analysis comparing post-operative CSF leak rate in terms of reconstruction methods
| Reconstruction method | n | CSF leak n | CSF leak % | p-value |
|---|---|---|---|---|
| NSF | 331 | 31 | 9.4 | 0.273 |
| No NSF | 65 | 9 | 13.8 | |
| Glue/haemostat | 234 | 28 | 12.0 | 0.139 |
| No glue/haemostat | 162 | 12 | 7.4 | |
| Pack | 230 | 25 | 10.9 | 0.550 |
| No pack | 166 | 15 | 9.0 | |
| All | 396 | 40 | 10.1 | – |
P-values are calculated from chi-squared (χ2) statistical tests
Discussion
Since the advent of endoscopic endonasal skull base surgery, multiple methods for reconstruction of the cranial base have been described. Generally, there are four aspects that surgeons may consider using within a multi-layer reconstruction: graft repair of the intra-cranial defect, utilisation of a pedicled NSF, utilisation of tissue adhesive glue or haemostatic agents, and packing of the nasal cavity. We found that whilst the majority of studies evaluated in our study initiated their reconstruction with an intra-cranial graft. The material utilised was varied and ranged from autologous fat, fascia or cartilage, to synthetic collagen matrix with or without sealant systems such as DuraSeal® [Integra LifeSciences Corporation, Plainsboro, NJ].
There was also heterogeneity in the use of tissue glue, haemostatic agents, and packing within the nasal cavity. Hadad et al. first described the use of a vascularised pedicle NSF for skull base reconstruction following endoscopic endonasal approaches, demonstrating that addition of this to a multi-layer reconstruction could reduce post-operative CSF leak rate to just 5% [7]. However we found that a NSF was used in only 70.8% of studies, with a multi-layer reconstruction being employed in 75.0% of studies.
Our systematic review found an overall post-operative CSF leak rate of 10.1%. This is comparable with previous literature comparing open and endoscopic management of clival chordomas, where open transcranial CSF leak rate ranged from 9.5%—10.7% and endoscopic 5.0 – 10.3% [33, 34]. Whilst not reaching statistical significance, there was a higher rate of CSF leak in cases that did not adopt a NSF in their reconstruction (9.4% vs. 13.8%).
This lack of statistical significance may be explained by the inclusion of a number of studies where the overall incidence of CSF leak was somewhat higher than the average of 10.1%, ranging from 14–33% [16, 17, 23, 28, 29]. Dehdashti et al. reported a case series of 12 patients undergoing endoscopic resection of a clival chordoma, where four patients had a post-operative CSF leak (33%). They used a multi-layer reconstruction technique in the entire cohort, but adopted a NSF in the final five patients. The CSF leaks they encountered occurred in patients where a NSF was not utilised, although this was hypothesised to be secondary to the underlying patient anatomy, with the presence of large pre-mesencephalic and pre-pontine cisterns [29].
Rahme et al. reported results from a series of 23 endoscopic endonasal chordoma resections where a NSF was utilised for reconstruction in all cases [14]. There were six CSF leaks in total (26%). They found that patients who developed a CSF leak had a significantly higher mean tumour volume than those that did not and were more likely to have had intra-dural extension at presentation. Three of these were in cases of resection of a recurrent tumour. These findings would suggest that tumour size, level of invasiveness, and technicality of dissection due to scar tissue formation are all potentially important factors in predicting risk of a CSF leak.
These included studies with higher observed leak rates are likely to have skewed our overall average in favour of higher CSF leaks and therefore influenced the lack of statistical significance observed when comparing the use of an NSF to without. Furthermore, in all 11 individual studies where the CSF leak rate was greater than our mean of 10.9%, the rate of other observed complications in these studies accounted for 52.4% of the overall post-operative complication rate (excluding CSF leaks). This suggests that in studies where a higher leak rate was observed, there was an overall higher complication rate, perhaps related to tumour complexity and/or surgeon experience rather than reconstructive methods.
Tissue glue and nasal packing are utilised by some centres in an attempt to create a seal at the edge of the graft material and to provide additional support to the reconstruction. The foley catheter balloon can be utilised to tamponade graft material against the defect, counteracting gravity and the pulsations of the brain. We found however that there was actually a higher non-statistically significant rate of CSF leak in the studies where glue/haemostat (12.0% vs. 7.4%, p = 0.139) and/or nasal packs (10.9% vs. 9.0%, p = 0.550) were utilised. This is somewhat contradictory to what one might expect but may be related to the multitude of both surgeon and tumour-specific factors involved in determining occurrence of CSF leak as described previously. In addition, there may be large variance in how glue is used. There may be misuse for example being placed between layers which may contribute to greater leak rates compared to no glue. Furthermore, we have assumed nasal packs were not used unless explicitly stated given the methodology of our review. There were also no studies that differentiated between absorbable and non-absorbable packs. There may have been studies using packs without documenting them, which may contribute to the lack of observed statistical significance.
Approximately one third of the included studies provided details regarding the use of a LD. Authors varied in their approach with some preferring to pre-emptively insert a LD either in all patients or based on intra-operative findings, whilst others adopted LDs only in the case of a confirmed post-operative CSF leak. However, in the studies where LDs were utilised following a pre-emptive approach, they did not provide specific data on patient drain status in relation to post-operative CSF leaks that they encountered. We were therefore unable to undertake any subgroup analysis on this particular subset of patients.
There are several limitations of this systematic review that should be kept in consideration. In terms of the quality of the evidence available, this was limited to case series and case reports and thus is a synthesis of level 3 and 4 evidence only. It therefore will represent data from a variety of treatment approaches, adjuvant therapy regimens and patient characteristics. These original reports may also be subject to selection and publication biases. Our search was also limited to the English language only, there may therefore be some degree of linguistic bias in our systematic review.
We found there was significant variation between studies in terms of the data reported and the level of detail included. Multiple studies did not report data on demographic points such as patient gender and age, or operative outcomes such as GTR. A number of studies also lacked a specific and comprehensive breakdown of their reconstruction method and its justification, and even fewer disclosed complete data relating to the use of LDs. This limited our ability to include these variables within our subgroup analysis of CSF leak rates.
In addition, there were many other cofounders that would affect CSF leak rates. Such examples include patient age, tumour size and extent, surgeon experience and any adjuvant treatment. While we evaluated and compared many baseline characteristics between studies (Table 2), we were not able to account for many other cofounders as described. Indeed, even within the baseline charactersitcs we have reported, there is heterogeneity with respect to patient demographics such as age and tumour factors such as GTR. This likely impacts on our overall effect estimate of 10.1% as an overall CSF leak rate. Despite this, we have synthesised all the available evidence pertaining to the endoscopic resection of CC. In addition, our value of 10.1% is in keeping with other reported estimates of CSF leaks in this patient cohort [33, 34].
As with all endoscopic skull base procedures, a variety of materials and repair strategies can be utilised and vary based on surgeon preference. As eluded to previously, most studies did not provide detailed accounts for the individual approach taken but it was clear in most cases, a multi-layered reconstruction was best suited to large clival defect repairs. The NSF has revolutionised skull base repair, especially in endoscopic clival approaches, however it may not necessarily be indicated in the absence of intra-operative CSF leaks. Again, it was not always clear which cases had intra-operative leaks and which did not, and whether a tailored approach to reconstruction was utilised on this basis. The length of patient follow-up also ranged significantly (6—68 months), and in multiple studies details on follow up were omitted.
Conclusion
In conclusion, our data is the most up-to-date synthesis of the existing literature surrounding the EEA to CCs with respect to reconstruction and post-operative CSF leaks. It demonstrates that EEA to CCs provides comparable outcomes to open approaches, with most authors employing a multi-layered reconstruction method. Although not statistically significant, the majority of authors prefer to use a NSF due to improved outcomes from repair, especially in higher flow site areas [35].
To date, the evidence does not demonstrate any statistically significant difference in CSF leak rate between all reconstruction methods used. However, this is likely due to a variety of surgeon and patient factors such as tumour size and invasiveness. Our results were somewhat limited by the reporting of reconstruction in the available literature. Higher quality prospective randomised multi-centric studies, with reporting of specific repair techniques need to be conducted to be enable more valid subgroup analyses in order to draw statistically significant conclusions. This will enable future systematic reviews and analyses to provide a more accurate consensus regarding optimal methods of reconstruction in this field.
Acknowledgements
We would like to acknowledge Jennifer Manders of the University Hospitals Birmingham (UHB) Library and Knowledge Service who helped design our search which was conducted on 1st March 2021.
Data Availability Statement
Data available from included articles as referenced.
Declarations
Conflicts of Interest
The authors report no conflicts of interest. All authors can confirm that they have no financial or non-financial interests that are directly or indirectly related to the submitted work to declare.
Human and Animal Rights
There was no research involving human participants or animals pertaining to this production of this manuscript.
Informed Consent
There were no requirements for informed consent to be sought with respect to the production of this manuscript.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Maio SD, Kong E, Yip S, et al. Converging paths to progress for skull base chordoma: review of current therapy and future molecular targets. Surg Neurol Int. 2013;4:72. doi: 10.4103/2152-7806.112822. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Smoll NR, Gautschi OP, Radovanovic I, et al. Incidence and relative survival of chordomas: the standardized mortality ratio and the impact of chordomas on a population. Cancer. 2013;119(11):2029–2037. doi: 10.1002/cncr.28032. [DOI] [PubMed] [Google Scholar]
- 3.Zou Y, Neale N, Sun J, et al. Prognostic factors in clival chordomas: an integrated analysis of 347 patients. World Neurosurg. 2018;118:e375–e387. doi: 10.1016/j.wneu.2018.06.194. [DOI] [PubMed] [Google Scholar]
- 4.Houessinon A, Boone M, Constans JM, et al. Sustained response of a clivus chordoma to erlotinib after imatinib failure. Case Rep Oncol. 2015;8:25–29. doi: 10.1159/000371843. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wang L, Tian K, Wang K, et al. Factors for tumor progression in patients with skull base chordoma. Cancer Med. 2016;5(9):2368–2377. doi: 10.1002/cam4.834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zoli M, Milanese L, Bonfatti R, et al. Clival chordomas: considerations after 16 years of endoscopic endonasal surgery. J Neurosurg. 2018;128(2):329–338. doi: 10.3171/2016.11.JNS162082. [DOI] [PubMed] [Google Scholar]
- 7.Hadad G, Bassagasteguy L, Carrau RL, et al. A novel reconstructive technique after endoscopic expanded endonasal approaches: vascular pedicle nasoseptal flap. Laryngoscope. 2006;116(10):1882–1886. doi: 10.1097/01.mlg.0000234933.37779.e4. [DOI] [PubMed] [Google Scholar]
- 8.Yasuda M, Bresson D, Chibbaro S, et al. Chordomas of the skull base and cervical spine: clinical outcomes associated with a multimodal surgical resection combined with proton-beam radiation in 40 patients. Neurosurg Rev. 2012;35(2):171–183. doi: 10.1007/s10143-011-0334-5. [DOI] [PubMed] [Google Scholar]
- 9.Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group Reprint–preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Phys Ther. 2009;89(9):873–880. doi: 10.1093/ptj/89.9.873. [DOI] [PubMed] [Google Scholar]
- 10.Oishi Y, Tamura R, Takahashi S, et al. A comparative study between traditional microscopic surgeries and endoscopic endonasal surgery for skull base chordomas. World Neurosurg. 2020;134:e1099–e1107. doi: 10.1016/j.wneu.2019.11.113. [DOI] [PubMed] [Google Scholar]
- 11.Wang Q, Wang Y, Wang J, Wang Y. Clinical classification of clival chordomas for transnasal approaches. Neurosurg Rev. 2020;43(4):1201–1210. doi: 10.1007/s10143-019-01153-w. [DOI] [PubMed] [Google Scholar]
- 12.Soloperto D, Fabbris C, De Rossi S, Musumeci A, Marchioni D. Endoscopic endonasal surgery of clival chordomas: preliminary results. Indian J Otolaryngol Head Neck Surg. 2019;71(4):486–491. doi: 10.1007/s12070-019-01661-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yousaf J, Afshari FT, Ahmed SK, Chavda SV, Sanghera P, Paluzzi A. Endoscopic endonasal surgery for Clival Chordomas - a single institution experience and short term outcomes. Br J Neurosurg. 2019;33(4):388–393. doi: 10.1080/02688697.2019.1567683. [DOI] [PubMed] [Google Scholar]
- 14.Rahme RJ, Arnaout OM, Sanusi OR, Kesavabhotla K, Chandler JP. Endoscopic approach to clival chordomas: The Northwestern experience. World Neurosurg. 2018;110:e231–e238. doi: 10.1016/j.wneu.2017.10.146. [DOI] [PubMed] [Google Scholar]
- 15.Shimony N, Gonen L, Shofty B, Abergel A, Fliss DM, Margalit N. Surgical resection of skull-base chordomas: experience in case selection for surgical approach according to anatomical compartments and review of the literature. Acta Neurochir (Wien) 2017;159(10):1835–1845. doi: 10.1007/s00701-016-3032-9. [DOI] [PubMed] [Google Scholar]
- 16.Catapano G, de Notaris M, Di Maria D, et al. The use of a three-dimensional endoscope for different skull base tumors: results of a preliminary extended endonasal surgical series. Acta Neurochir (Wien) 2016;158(8):1605–1616. doi: 10.1007/s00701-016-2847-8. [DOI] [PubMed] [Google Scholar]
- 17.Garzaro M, Zenga F, Raimondo L, et al. Three-dimensional endoscopy in transnasal transsphenoidal approach to clival chordomas. Head Neck. 2016;38(Suppl 1):E1814–E1819. doi: 10.1002/hed.24324. [DOI] [PubMed] [Google Scholar]
- 18.Chibbaro S, Cornelius JF, Froelich S, et al. Endoscopic endonasal approach in the management of skull base chordomas–clinical experience on a large series, technique, outcome, and pitfalls. Neurosurg Rev. 2014;37(2):217–225. doi: 10.1007/s10143-013-0503-9. [DOI] [PubMed] [Google Scholar]
- 19.Iacoangeli M, Di Rienzo A, di Somma LG, et al. Improving the endoscopic endonasal transclival approach: the importance of a precise layer by layer reconstruction. Br J Neurosurg. 2014;28(2):241–246. doi: 10.3109/02688697.2013.835375. [DOI] [PubMed] [Google Scholar]
- 20.Ishii Y, Tahara S, Teramoto A, Morita A. Endoscopic endonasal skull base surgery: advantages, limitations, and our techniques to overcome cerebrospinal fluid leakage: technical note. Neurol Med Chir (Tokyo) 2014;54(12):983–990. doi: 10.2176/nmc.st.2014-0081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shidoh S, Toda M, Kawase T, et al. Transoral vs. endoscopic endonasal approach for clival/upper cervical chordoma. Neurol Med Chir (Tokyo) 2014;54(12):991–998. doi: 10.2176/nmc.st.2014-0135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Saito K, Toda M, Tomita T, Ogawa K, Yoshida K. Surgical results of an endoscopic endonasal approach for clival chordomas. Acta Neurochir (Wien) 2012;154(5):879–886. doi: 10.1007/s00701-012-1317-1. [DOI] [PubMed] [Google Scholar]
- 23.Tan NC, Naidoo Y, Oue S, et al. Endoscopic surgery of skull base chordomas. J Neurol Surg B Skull Base. 2012;73(6):379–386. doi: 10.1055/s-0032-1321508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Taniguchi M, Kohmura E. Endoscopic endonasal removal of laterally extended clival chordoma using side-viewing scopes. Acta Neurochir (Wien) 2012;154(4):627–632. doi: 10.1007/s00701-011-1225-9. [DOI] [PubMed] [Google Scholar]
- 25.Fraser JF, Nyquist GG, Moore N, Anand VK, Schwartz TH. Endoscopic endonasal minimal access approach to the clivus: case series and technical nuances. Neurosurgery. 2010;67(3 Suppl Operative):ons150–ons158. doi: 10.1227/01.NEU.0000383130.80179.41. [DOI] [PubMed] [Google Scholar]
- 26.Holzmann D, Reisch R, Krayenbühl N, Hug E, Bernays RL. The transnasal transclival approach for clivus chordoma. Minim Invasive Neurosurg. 2010;53(5–6):211–217. doi: 10.1055/s-0030-1267929. [DOI] [PubMed] [Google Scholar]
- 27.Arbolay OL, González JG, González RH, Gálvez YH. Extended endoscopic endonasal approach to the skull base. Minim Invasive Neurosurg. 2009;52(3):114–118. doi: 10.1055/s-0028-1119414. [DOI] [PubMed] [Google Scholar]
- 28.Stippler M, Gardner PA, Snyderman CH, Carrau RL, Prevedello DM, Kassam AB. Endoscopic endonasal approach for clival chordomas. Neurosurgery. 2009;64(2):268–278. doi: 10.1227/01.NEU.0000338071.01241.E2. [DOI] [PubMed] [Google Scholar]
- 29.Dehdashti AR, Karabatsou K, Ganna A, Witterick I, Gentili F. Expanded endoscopic endonasal approach for treatment of clival chordomas: early results in 12 patients. Neurosurgery. 2008;63(2):299–309. doi: 10.1227/01.NEU.0000316414.20247.32. [DOI] [PubMed] [Google Scholar]
- 30.Hwang PY, Ho CL. Neuronavigation using an image-guided endoscopic transnasal-sphenoethmoidal approach to clival chordomas. Neurosurgery. 2007;61(5 Suppl 2):212–218. doi: 10.1227/01.neu.0000303219.55393.fe. [DOI] [PubMed] [Google Scholar]
- 31.Frank G, Sciarretta V, Calbucci F, Farneti G, Mazzatenta D, Pasquini E. The endoscopic transnasal transsphenoidal approach for the treatment of cranial base chordomas and chondrosarcomas. Neurosurgery. 2006;59(1 Suppl 1):ONS50–ONS57. doi: 10.1227/01.NEU.0000219914.17221.55. [DOI] [PubMed] [Google Scholar]
- 32.Solares CA, Fakhri S, Batra PS, Lee J, Lanza DC. Transnasal endoscopic resection of lesions of the clivus: a preliminary report. Laryngoscope. 2005;115(11):1917–1922. doi: 10.1097/01.mlg.0000172070.93173.92. [DOI] [PubMed] [Google Scholar]
- 33.Komotar RJ, Starke RM, Raper DM, Anand VK, Schwartz TH. Endoscopic skull base surgery: a comprehensive comparison with open transcranial approaches. Br J Neurosurg. 2012;26(5):637–648. doi: 10.3109/02688697.2012.654837. [DOI] [PubMed] [Google Scholar]
- 34.Cannizzaro D, Tropeano MP, Milani D et al (2020) Microsurgical versus endoscopic trans-sphenoidal approaches for clivus chordoma: a pooled and meta-analysis [published online ahead of print, 2020 May 29]. Neurosurg Rev. 10. 10.1007/s10143-020-01318-y [DOI] [PubMed]
- 35.Harvey RJ, Parmar P, Sacks R, Zanation AM. Endoscopic skull base reconstruction of large dural defects: a systematic review of published evidence. Laryngoscope. 2012;122(2):452–459. doi: 10.1002/lary.22475. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data available from included articles as referenced.