Long-Term Results after Extra-Cranial Skull Base Reconstruction - A Cohort Study

  

    
Patient’s ID
    
Localization of mucocele
    
Approach for rhinobase    reconstruction
  
  

    
#7
    
Bilateral frontal sinus
    
External bilateral
  
  

    
#20
    
Bilateral sphenoid sinus
    
Endonasal Bilateral
  
  

    
#29
    
Right frontal sinus
    
External bilateral
  
  

    
#35
    
Right frontal sinus
    
External right
  
  

    
#43
    
Right frontal sinus
    
External right
  
  

    
#46
    
Left frontal sinus and ethmoid
    
External left
  
  

    
#69
    
Left sphenoid sinus
    
External bilateral
  
  

    
#62
    
n.i.
    
Endonasal right, external left
  
  

    
#1
    
n.i.
    
External left
  
  

    
#77
    
Left frontal sinus
    
External left, endonasal right
  
  

    
#101
    
Right frontal sinus
    
External bilateral
  
  

    
#124
    
Left frontal sinus
    
External left
  
  

    
#133
    
n.i.
    
External bilateral
  

Table 6: Clinical findings of those 13 patients who developed a mucocele. The
localization of the mucocele could not be revealed (n.i.) in three patients.

As indicated in the methods section of this manuscript, indication for skull base reconstruction was seen when there was either one or multiple of the above mentioned inclusion conditions.

Eight patients demonstrated rhinoliquorrhoea pre operatively. Sixty-four (58.2%) of 110 patients with available CT-scan reports, exhibited intracranial air in the CT scan, preoperatively. Forty-six (41.8%) of 110 patients had fracture lines without intracranial air in the CT scan. Two patients (#1 and #110) were operated because of meningitis.

Intraoperative findings: Focusing on all 138 patients, in 19 patients CSF- leakage was found intraoperatively. Among them are two patients who presented rhinoliquorrhoea preoperatively. Six patients demonstrated intracranial air and twelve patients demonstrated fracture lines without intracranial air in the pre-operative CT-scan. Localization of the fracture lines in those nineteen patients with intraoperatively detected CSF leakage is summarized in Table 3.

Pre-operatively, eight patients demonstrated rhino-liquorrhoea as indicated above. Laceration of the dura with CSF leakage could be confirmed intraoperatively in two of these patients. The remaining six patients showed fractures without intraoperative CSF leakage.

Meningitis: Two patients (#1, #110) presented with rhinobase fracture and meningitis. These fractures were closed. Patient #1 received rhinobase exploration in a different hospital where rhinoliquorrhoea was observed intraoperatively. The patient developed meningitis and persisting rhino- liquorrhoea. The patient was transferred to our department where we performed rhinobase reconstruction. Patient #110 had rhinobase reconstruction in another hospital because of rhinobase fracture. One month later the patient was referred to our department with meningitis and encephalitis. Intraoperatively, the fracture lines were explored but no active CSF leakage was found. The fracture lines were covered by TachoSil^® compare (Table 4).

Reconstruction materials: Seven different materials or a combination of these were used for the reconstruction of rhinobase compare (Table 5,6). Most frequently TachoSil^® and (bovine or porcine) pericardial patch were applied. The rhinobase was reconstructed with a combination of two materials in 39 patients. Three different materials were applied in 6 patients.

5.3. Postoperative complications

Rhinoliquorrhoea: Five patients (#90, #46, #19, #134, #133) developed rhinoliquorrhoea postoperatively. Indication for rhinobase reconstruction in those 5 patients was set as following: One patient demonstrated rhinoliquorrhoea pre-operatively. In two patients CSF- leakage was found intraoperatively. The remaining two patients had fracture lines without intraoperative CSF- leakage. None of these five patients had intracranial air in the preoperative CT scan. All patients received revision surgery where active CSF-leakage could be visualized. Neither lumbar drainages nor neurosurgical transcranial approaches were needed to successfully seal these CSF-fistulas.

5.4. Meningitis following surgery in our hospital

Three patients (#90, #123, #133) developed meningitis following surgery in our department. One of these patients (#90) developed meningitis 6 years after closed rhinobase fracture and rhinobase reconstruction. In the initial surgery there was freely exposed dura mater without bony coverage and no signs of CSF-leakage. When the patient presented with meningitis Fluorescein was applied intrathecally. Intraoperative exploration of the rhinobase did not reveal rhinoliquorrhoea. The previous fracture area was again covered with TachoSil^®. Under intravenous antibiotics the patient recovered. The second patient (#123) presented with meningitis 3 weeks after closed rhinobase reconstruction following massive rhinobase fracture with discharge of brain tissue into the nasal cavity. The patient received revision surgery for rhinobase exploration but no signs of CSF- leakage could be found. The third patient (#133) was initially operated because of a rhinobase fracture in the frontal sinus and in the sphenoid sinus. There were neither preoperative nor intraoperative signs of CSF-leakage. The initial CT scan showed no intracranial air. At the 5^th post-operative day rhinoliquorrhoea was observed and the patient received revision surgery. The patient was admitted to the hospital 14 days after revision surgery with meningitis. Again, rhinobase was explored. Purulent secretion was found in the frontal sinus. There was no sign of rhinoliquorrhoea.

5.5. Development of mucoceles

Thirteen patients developed a mucocele of the paranasal sinuses within the follow up period. Mucoceles developed in 2 out of 101 (2%) transnasal endoscopically treated sides while mucoceles were observed in 11 out of 92 (12%) sides which were operated using an external approach. The mucocele could not be assigned to the used approach in two additional patients. Qui-square-test revealed significant less mucoceles in the endoscopically treated patients (qui-square-value 7.63; df=1; p=0.006).

In detail: One patient developed a mucocele in the left and the right sphenoid sinus following a bilateral endoscopic rhinobase reconstruction. 10 patients developed unilateral mucoceles on the side that was operated via an external approach. Patient (#133) received a bilateral external approach but could not indicate the side of the mucocele that was operated in different hospital. Patient (#1) who received rhinobase reconstruction via an external approach on the left side previously, indicated surgery for a mucocele in a different hospital without remembering the side. A second patient (#62) who was operated via a transnasal approach on the right side and via an external approach on the left side indicated surgery for a mucocele elsewhere but could not remember the side.

Discussion

The present study sheds new light on the outcome of rhinobase reconstruction in patients who were treated surgically because of rhinobase fractures. Among the studied population 1.4%, 5.7%, 58.2% and 100% of the patients were operated because of meningitis, rhinoliquorrhoea, intracranial air and fracture lines in the CT scans, respectively. Due to the diversity in the management of skull base fractures, the majority of our patients would not have been operated in other centers [9]. However, Elies justified rhinobase exploration in cases of CT-morphological fractures [12]. The author indicated that in 25% of his patients a CSF leakage was found intraoperatively although there was no clinical rhinoliquorrhoea preoperatively. Nowadays, it is widely believed that CSF leakages should be closed surgically to avoid ascending infections like meningitis although no study specifically demonstrated a beneficial effect, yet [2,7,8,10]. Eljamel reports a rate of meningitis in cases of non-surgically treated rhinoliquorrhoea of 6.6% and 7.6% within one week and one year, respectively [13]. Other study groups report much higher rates of meningitis of 29% or even 85% within 5 years after the trauma [2,3]. Bernal-Sprekelsen and Co-authors state that this high incidence of meningitis is not acceptable and favor early endoscopic closure of the fistula. Preoperative clinical manifest rhinoliquorrhoea was found in 8 (5.7%) of our patients. Interestingly, in only 2 out of these 8 patients active CSF leakage could be visualized while exploring the rhinobase in those patients. On the other hand focusing on the patients who preoperatively did not demonstrate rhinoliquorrhoea, our data suggest that 19 of 130 patients (14.6%) had active CSF leakage intraoperatively. These facts illustrate the dilemma. On one side we could visualize the CSF leakage intraoperatively only in 25% of the patients who presented with rhinoliquorrhoea preoperatively. On the other side we found active CSF leakage intraoperatively in 14.6% of the patients who did not present rhinoliquorrhoea preoperatively. There are several explanations for these conditions. An initial fistula can be closed by e.g. a clot, a brain herniation or an early scar formation that make it impossible to visualize the fistula intraoperatively. On the other hand a fistula might be present but result in a non-clinically detectable rhinoliquorrhoea or the fistula opens when the intracranial pressure decreases while the patient is operated in general anesthesia [14,15]. One could assume that in almost 60% of the patients there could be a skull base fistula when extrapolating our finding that the rate of CSF leakage is four times higher than it can be visualized intraoperatively (CSF leakage could be visualized in 25% of the patients with preoperatively manifest rhinoliquorrhoea and CSF leakage could be found intraoperatively in 14.6% of the patients without preoperatively manifest rhinoliquorrhoea). That would give good reason for rhinobase reconstruction in patients with CT-morphologic skull base fractures even in cases without apparent CSF leakage. Another approach could be to identify those patients who are on high risk to present rhinoliquorrhoea intraoperatively.

In the present study the rate of surgically detected active CSF leakage was 50%, 25%, 11% and 26% in those with meningitis, pre-operative rhinoliquorrhoea, intracranial air and fracture lines, respectively. Table 3 illustrates the diverse localizations of the CSF leakages at the rhinobase. It appears that our pre-operative indication for surgery as well as the localization of the fracture lines cannot predict the appearance of intraoperative CSF leakage. However, the motivation for surgery in our studied patients was to reduce their risk to develop meningitis. The above mentioned incidence of meningitis (6.6 to 85%) was calculated in patients with clinically apparent rhinoliquorrhoea [2,3,13]. It is not clear whether patients who have no preoperative rhinoliquorrhoea but present an intraoperative CSF leakage have the same risk to develop meningitis like patients with preoperative clinically apparent rhinoliquorrhoea. According to a study from the 1950´s none of 50 patients with skull base facture without rhinoliquorrhoea developed meningitis within 5 years (range 1.5 to 12 years) [5]. That would mean that 94% of our studied population received over-treatment. Other studies reported an accumulative risk of meningitis of 85% after 10 years with a mortality rate of 25 - 50% [3]. In harsh contrast to later studies, the same study indicates that none of the 27 patients who received conservative treatment of rhinoliquorrhoea developed meningitis during the above mentioned follow-up period. Another study from the pre-World War II era indicated 6 cases (24%) of meningitis following skull base fracture without rhinoliquorrhoea and x-ray morphological intracranial air [6]. Four of these 6 patients died as a result of intracranial infections. Despite our efforts to prevent meningitis, three patients developed meningitis (2.2%) following rhinobase reconstruction in our department. Interestingly, none of the 3 patients presented active CSF leakage prior or during revision surgery. Only in one patient a focus with purulent secretion was found close to the rhinobase that could explain ascending meningitis. The mechanistic concept of direct contact of pus with the dura mater that results in bacterial meningitis is certainly just one possible explanation. Lung infections and endocarditis are predisposing factors for the development of bacterial meningitis which suggests a hematogenous spread of the germs [18,19].

The rate of 2.2% of post-operative meningitis in the present study is in line with previous studies from Bernal-Sprekelsen et al. and from Rocchi G et al. which report rates of 2.6% and 5.6% [20,21]. Bernal-Sprekelsen and colleagues studied rates of post-operative meningitis following transnasal endoscopic closure of dura fistula in 39 patients with proven rhinoliquorrhoea or meningitis. The authors indicate a post-operative meningitis rate of 2.6% within a follow-up period of 65 month (min. 22 max. 120 month). At first sight their meningitis rate appears comparably low like in the present study (2.2% among 138 patients). But, when looking at our ten patients with pre-operative CSF-leakage or meningitis, who match with the studied patients of Bernal-Sprekelsen et al., we had not a single case of meningitis during the follow-up interval. None of the three patients who developed meningitis in the present study strictly fit the indicated inclusion criteria of Bernal-Sprekelsen´s and Co-authors´ study. Additionally, the authors did not follow-up those patients with rhinobase fracture without CSF-leakage. Doing though, we would have missed our three patients who developed meningitis.

Every medical intervention implicates the risk of side effects. Specifically, rhinobase reconstruction put the patient on risk to suffer surgery-related complications including bleeding, wound infections or endonasal scaring. The present study focused on mucoceles - one possible sequel. In the present study 13 patients developed mucoceles. At least 2 patients (maximum 4 patients) presented a bilateral mucocele. In other words, a mucocele occurred in 7.4% of the operated sides or in 9.4% of the patients. Causes for mucocele formation in the paranasal sinuses are divers. Besides sporadic development of mucoceles, non-iatrogenic risk factors are chronic rhinosinusitis and traumata e.g. fracture of the paranasal sinuses [22-24]. Interestingly, no single mucocele developed on the non-fractured side in those 72 cases with unilateral rhinobase fracture. Chobillon and Jankowki estimate the risk for mucocele development in non-surgically treated chronic rhinosinusitis at 0.6% while the figure increases to 5.3% within 6 years in transnasal endoscopically treated patients [24]. The rate of mucoceles in the present study was twice as high as in Chobillon and Jankowki´s study. This figure might be explained by the high rate of external approaches in the present study. Significantly (p=0.006) less mucoceles developed following endoscopic surgery compared with external approaches. In our patients, mucoceles developed in 2 out of 101 (2%) transnasal endoscopically treated sides while mucoceles were observed in 11 out of 92 (12%) sides which were operated using an external approach. That is a surprisingly low incidence in the endoscopically treated patients; especially considering studies indicating that mucoceles arise significantly earlier following endoscopic transnasal surgery compared to external approaches to the sinuses [25,26]. On one hand, rhinobase reconstruction was performed in non-inflamed sinuses in the present study. That fact might underline the role of chronic inflammation in the development of mucoceles [27]. On the other hand, 12% of mucoceles following external approaches illustrate a comparable high incidence. Elies reports mucocele formation in 22.4% of patients who received external approaches for rhinobase reconstruction [12]. That high incidence of mucoceles following external approaches should be avoided by favoring transnasal endoscopic surgery. However, the lateral recess of the frontal sinus is not assessable transnasally with rigid endoscopes that makes it impossible to close skull base defects in that location. Transorbital endoscopic surgery is an alternative approaches that might be less invasive for the lateral recess of the frontal sinus [28]. Nevertheless, future studies have to prove whether this approach might be associated with less long-term sequelae than extensive external approaches.

Conclusion

In 14% surgical rhinobase exploration unveils dura laceration with CSF-leakage in patients who demonstrate rhinobase fractures in the CT-scan and have no clinical signs of rhinoliquorrhoea. Although it is nowadays widely believed that in patients with rhinoliquorrhoea skull base reconstruction reduces the risk for post-operative meningitis, a prospective study is needed that systematically investigates that issue. Every physician who recommends skull base reconstruction in patients with rhinoliquorrhoea should consider rhinobase exploration even in patients with rhinobase fractures without rhinoliquorrhoea because a significant number of patients with dura laceration can just be identified surgically. Furthermore, transnasal endoscopic surgery should be favored for skull base reconstruction in order to reduce the incidence of mucoceles following external approaches.

All patients who participated in the study gave informed consent.

References

1. Stammberger H, Greistorfer K, Wolf G, Luxenberger W. [Surgical occlusion of cerebrospinal fistulas of the anterior skull base using intrathecal sodium fluorescein]. Laryngorhinootologie. 1997; 76: 595-607.
2. Bernal-Sprekelsen M, Bleda-Vazquez C, Carrau RL. Ascending meningitis secondary to traumatic cerebrospinal fluid leaks. Am J Rhinol. 2000; 14: 257-259.
3. Eljamel MS, Foy PM. Acute traumatic CSF fistulae: the risk of intracranial infection. Br J Neurosurg. 1990; 4: 381-385.
4. Calvert CA, Cairns H. Discussion on injuries of the frontal and ethmoid sinuses. Proc R Soc Med .1942; 35: 805-810.
5. Brawley BW, Kelly WA. Treatment of basal skull fractures with and without cerebrospinal fluid fistulae. J Neurosurg. 1967; 26: 57-61.
6. Discussion on Injuries of the Frontal and Ethmoidal Sinuses. Proc R Soc Med. 1942; 35: 805-810.
7. Hegazy HM, Carrau RL, Snyderman CH, Kassam A, Zweig J. Transnasal  endoscopic  repair  of cerebrospinal fluid rhinorrhea: a meta-analysis. Laryngoscope. 2000; 110: 1166-1172.
8. Bernal-Sprekelsen M, Alobid I, Mullol J, Trobat F, Tomas-Barberan M. Closure of cerebrospinal fluid leaks prevents ascending bacterial meningitis. Rhinology. 2005; 43: 277-281.
9. Piccirilli M, Anichini G, Cassoni A, Ramieri V, Valentini V, Santoro A, et al. Anterior cranial fossa traumas: clinical value, surgical indications, and results-a retrospective study on a series of 223 patients. J Neurol Surg B Skull Base. 2012; 73: 265-272.
10. Schoentgen C, Henaux PL, Godey B, Jegoux F. Management of post-traumatic cerebrospinal fluid (CSF) leak of anterior skull base: 10 years experience. Acta Otolaryngol.  2013; 133: 944-950.
11. Swinson BD, Jerjes W, Thompson G. Current practice in the management of frontal sinus fractures. J Laryngol Otol. 2004; 118: 927-932.
12. Elies W. [The present state of frontal skull-base surgery (author's transl)]. Laryngol Rhinol Otol (Stuttg). 1982; 61: 42-47.
13. Eljamel MS. Antibiotic prophylaxis in unrepaired CSF fistulae. Br J Neurosurg. 1993; 7: 501-505.
14. Nohra G, Jabbour P, Haddad A, Abouhamad W, Abilahoud G, Okais N, et al. [Subcranial subfrontal approach for the treatment of extensive cerebrospinal fluid leaks]. Neurochirurgie. 2002; 48: 87-91.
15. Hubbard JL, McDonald TJ, Pearson BW, Laws ER Jr. Spontaneous cerebrospinal fluid rhinorrhea: evolving concepts in diagnosis and surgical management based on the Mayo Clinic experience from 1970 through 1981. Neurosurgery. 1985; 16: 314-21.
16. Durand ML, Calderwood SB, Weber DJ, Miller SI, Southwick FS, Caviness VS Jr, et al. Acute bacterial meningitis in adults. A review of 493 episodes. N Engl J Med. 1993; 328: 21-28.
17. Schick B, Kahle G, Weber R, Draf W. [Experiences in diagnosis of occult traumatic dural lesions of the anterior cranial base]. Laryngorhinootologie. 1998; 77: 144-149.
18. Van de Beek D, de Gans J, Spanjaard L, Weisfelt M, Reitsma JB, Vermeulen M, et al. Clinical features and prognostic factors in adults with bacterial meningitis. N Engl J Med. 2004; 351: 1849-1859.
19. Brouwer MC, Keizerweerd GD, De Gans J, Spanjaard L, Van De Beek D: Community acquired Staphylococcus aureus meningitis in adults. Scand J Infect Dis. 2009; 41: 375-377.
20. Rocchi G, Caroli E, Belli E, Salvati M, Cimatti M, Delfini R, et al. Severe craniofacial fractures with frontobasal involvement and cerebrospinal fluid fistula: indications for surgical repair. Surg Neurol. 2005; 63: 559-563; discussion 563-554.
21. Bernal-Sprekelsen M, Alobid I, Mullol J, Trobat F, Tomás-Barberán M. Closure of cerebrospinal fluid leaks prevents ascending bacterial meningitis. Rhinology. 2005; 43: 277-81.
22. Koudstaal MJ, van der Wal KG, Bijvoet HW, Vincent AJ, Poublon RM: Post-trauma mucocele formation in the frontal sinus; a rationale of follow-up. Int J Oral Maxillofac Surg. 2004; 33: 751-754.
23. Lund VJ: Anatomical considerations in the aetiology of fronto-ethmoidal mucoceles. Rhinology. 1987; 25: 83-88.
24. Chobillon MA, Jankowski R. Relationship between mucoceles, nasal polyposis and nasalisation. Rhinology, 2004; 42: 219-224.
25. Rombaux P, Bertrand B, Eloy P, Collet S, Daele J, Bachert C, Claes J, et al. Endoscopic endonasal surgery for paranasal sinus mucoceles. Acta Otorhinolaryngol Belg. 2000; 54: 115-122.
26. Raynal M, Peynegre R, Beautru R, Coste A. [Sinus mucoceles and surgery in iatrogenic diseases]. Ann Otolaryngol Chir Cervicofac. 1999; 116: 85-91.
27. Lund VJ, Harvey W, Meghji S, Harris M. Prostaglandin synthesis in the pathogenesis of fronto-ethmoidal mucoceles. Acta Otolaryngol. 1988; 106: 145-151.
28. Moe KS, Kim LJ, Bergeron CM. Transorbital endoscopic repair of cerebrospinal fluid leaks. Laryngoscope. 2011; 121: 13-30.