CLINICAL STUDY
Intracranial Repair of Posttraumatic Cerebrospinal Fluid Rhinorrhea Associated With Recurrent Meningitis Can Yaldiz, MD,* Nail Ozdemir, MD,† Onur Yaman, MD,‡ İsmail Ertan Seyin, MD,‡ and Serdar Oguzoglu, MD‡ Objective: The purposes of this study are to assess the efficacy of our intracranial surgery and evaluate the association between failure after first surgical repair and the risk factors that have been applied on a group of 13 patients affected by posttraumatic cerebrospinal fluid rhinorrhea associated with recurrent meningitis. Methods: We retrospectively collected data on 13 patients referred to our institution. All patients had history of head trauma and experienced 2 or more episodes of meningitis. Results: Three of the 13 patients had craniectomy defect due to previous trauma and surgery, 9 patients had linear fracture, and 1 patient had no apparent fracture line on preoperative radiologic evaluation. Ten of the 13 patients had identified frontal bone fracture involving the frontal sinus during surgery. Dural tear was identified intradurally and was repaired using a fascia lata graft with or without fibrin glue. Fibrin glue was applied over the suture in 7 patients. Three of the 13 patients had large dural defects. Conclusions: The size of bone and dural defect seems to be an important prognostic factor of episodes of meningitis. The use of fibrin glue to fixate fascia lata graft did not benefit the outcome. Key Words: Cerebrospinal fluid rhinorrhea, head trauma, meningitis, surgical repair (J Craniofac Surg 2015;26: 170–173)
P
osttraumatic cerebrospinal fluid (CSF) rhinorrhea is caused by tearing of the arachnoid and dura mater associated with skull fractures. Approximately 80% of all cases of CSF rhinorrhea occur after head injury, and this injury commonly occurs at the floor of the anterior cranial fossa. The CSF rhinorrhea implies leakage of CSF through the meninges and skull base into the paranasal sinuses with subsequent leakage through the anterior nasal apertures.1,2 The incidence of meningitis from CSF rhinorrhea is reported to be 10% and up to 40% during long-term follow-up.3 The mortality rate of
From the *Department of Neurosurgery, Sakarya University Training and Research Hospital, Sakarya; †Department of Neurosurgery, İzmir Tepecik Training and Research Hospital, İzmir; and ‡Department of Neurosurgery, İzmir Atatürk Training and Research Hospital, İzmir, Turkey. Received March 16, 2014. Accepted for publication June 20, 2014. Address correspondence and reprint requests to Can Yaldiz, MD, Department of Neurosurgery, Sakarya University Training and Research Hospital, Şirinevler Mh., 54100 Sakarya, Turkey; E-mail: [email protected] Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.jcraniofacialsurgery.com). The authors report no conflicts of interest. Copyright © 2014 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000001181
170
posttraumatic CSF rhinorrhea associated with meningitis is approximately 10%.2,4 Meningitis may develop acutely or even decades after the trauma.5 Because of the attendant risk for meningitis, brain abscess, and pneumocephalus, all persistent CSF rhinorrhea should be repaired.6 The purposes of this retrospective analysis are to determine the relationship between the fracture type, bone defect, dural defect, pneumocephalus, brain abscess, precise preoperative localization, fascia lata graft, fibrin glue, and persistent CSF rhinorrhea and assess the clinical outcomes and prognostic factors of patients admitted to our clinic with posttraumatic CSF rhinorrhea associated with recurrent meningitis who were managed by surgical intracranial repair.
MATERIALS AND METHODS Thirteen patients with clinical CSF rhinorrhea were included in this study at the Neurosurgery Clinic of İzmir Atatürk Research and Training Hospital from September 2003 to June 2009. All patients had cranial trauma and episodes of meningitis before our study. None of the patients had clinically detectable meningitis at admission. Nonetheless, each patient had received antibiotic therapy during episodes of meningitis. Retrospective analysis of data including demographic details, etiology, location of CSF leak, imaging results, surgical techniques, complications, recurrence, and surgical outcomes were examined. Indications for surgery included chronic occult rhinorrhea and recurrent meningitis. Iatrogenic leaks from previous neurosurgical or rhinosurgical surgery (such as removal of a pituitary adenoma, paranasal sinus surgery, endonasal procedure, orbital decompression) and spontaneous causes (such as congenital malformations or tumors) were excluded.
RESULTS Patient Data During the studied period, 13 patients were referred with apparent CSF rhinorrhea and recurrent meningitis due to chronic rhinorrhea. The patients consisted of 12 males and 1 female, ranging in age from 11 to 74 years (mean age, 28.6 y); 10 patients were younger than 30 years. History of a significant trauma to the head and/or obvious findings at surgery consistent with a traumatic origin were included in this study. Of the total 13 patients, 9 had been treated nonsurgically, and 4 had undergone cranial surgery. Surgery was necessary in 4 patients who had skull base fractures accompanied by a penetrating injury or who had depressed cranial fractures associated with cerebral pathology. Eight patients had been injured in traffic accidents, 4 patients had fallen, and 1 patient experienced a gunshot wound. A total of 13 patients had occult CSF rhinorrhea at admission, and all these individuals presented with recurrent meningitis after the original trauma. We reviewed the previous medical records of 13 patients who were treated for bacterial meningitis due to CSF rhinorrhea. In these patients, the mean interval from head injury to recurrent meningitis was 36 months (minimum, 1 mo; maximum, 132 mo). All patients experiences 2 or more episodes of bacterial
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
meningitis. The number of patients with repeated episodes of meningitis in this study was as follows: 2 attacks, 5 of 13; 3 attacks, 3 of 13; and 5 attacks, 3 of 13. Two patients in our series developed cerebral abscesses after meningitis, and all had been treated succesfully. One patient with cerebral abscess experienced a gunshot wound.
Preoperative Evaluation The mean Glasgow Outcome Scale score at admission was 4.61 in all patients. The Glasgow Outcome Scale was favorable for all 13 of the patients; none of the patients were severely disabled. A total of 6 patients had cranial nerve injuries (olfactory nerve, 2 patients; optic nerve, 4 patients). All patients underwent noncontrast high-resolution computed tomography (CT) with bone window images. Six patients required magnetic resonance imaging (MRI) to define the skull base defect. Three of the 13 patients had craniectomy defect and pneumocephalus due to previous trauma and surgery (including the 2 cerebral abscesses), and 9 patients had linear fracture. One patient had not identified fracture line or other pathology with CT and MRI (see Supplemental Digital Content, Table 1, http:// links.lww.com/SCS/A100).
Treatment A bicoronal scalp incision was made along the hairline. The flap is elevated anteriorly beyond the supraorbital ridges and glabella. Laterally, the dissection was done superficial to the temporalis fascia. The supraorbital nerves and vessels are identified and carefully dissected from the supraorbital notch. Wide opposure of the ethmoidal roof, the whole frontal region, the orbital roofs, and lateral walls up the orbit apex is performed. Sometimes, it was possible to remove frontal and orbital bone segments along fracture lines. Then, a bilateral frontal craniotomy and dural incision were made. With the aid of an operative microscope, gentle retraction of the frontal lobes was used to allow visualization of the bone and dural defects and the herniated brain tissue. Dural tear was identified intradurally and could not be closed primarily; it was repaired using a fascia lata graft under the operating microscope. Care was taken to try to obtain a water-tight closure. Fibrin glue was applied over the suture in 7 patients. Other 6 patients had used only fascia lata graft. Ten of the 13 patients had identified frontal bone fracture involving the frontal sinus. Three of the 13 patients had a dural defect of more than 1 cm. The anterior fossa was found to have numerous small 2- to 3-mm indentations, and approximately 2 cm behind the posterior wall of the frontal sinus, an outpouching of dura was found with a dural tear and herniation of brain tissue. The frontal sinuses are meticulously reconstructed. These patients were treated using mucozectomy and fat obliteration of the frontal sinuses associated with fascia lata duraplasty (see Supplemental Digital Content, Table 1, http://links.lww.com/SCS/A100). Lumbar drains were not used routinely in our patients. The mean postoperative stay in the intensive care unit was 48 hours, and the mean hospital stay was 26 days. Two patients had postoperative temporary nontension pneumocephalus. There were no meningitis, epidural or intradural hematoma, frontal lobe infarct or contusion, and seizures after surgery. The mean followup was 4.4 years, with a range of 6 months to 6 years. Success was achieved after the first surgery in 10 of the 13 patients (77%). In 3 patients who were reoperated on for recurrent rhinorrhea, a lumboperitoneal shunt was placed during the second surgery. Success was achieved in all 3 patients after a second surgery at repair (see Supplemental Digital Content, Table 1, http://links.lww. com/SCS/A100). In these cases of failure after first surgical repair, the risk factors evaluated for possible association included site and size defect, pneumocephalus, brain abscess, precise preoperative localization, and
CSF Rhinorrhea With Recurrent Meningitis
materials used for repair. Chi-squared test was used for statistical analysis. Association between failure after first surgical repair and the risk factors above was evaluated based on Fisher exact test. All calculated P values were 2-sided, and Ps < 0.05 were considered statistically significant. There were 2 cases of brain abscess after cranial trauma surgery in this reoperated group. Association between failure and presence of brain abscess was statistically significant (P < 0.038). There were 3 cases of a craniectomy defect and large dural defect (measuring more than 1 cm) in this reoperated group. Association between failure and presence of craniectomy defect was statistically significant (P < 0.038). Association between failure and large dural defect was statistically significant (P < 0.038). If the patient has had no history of craniotomy or brain abscess, rhinorrhea usually did not recur after the first operation. Intraoperative localization was confirmed by CT and/or MRI in only 5 patients. Pneumocephalus was related to an increased incidence of brain abscess. However, in patients with multiple sites of fracture, there was no significant increased number of recurrent meningitis. Fascia lata graft with fibrin glue was applied in 7 of the 13 patients. Three of the 13 patients required reoperation for recurrent leakage, and fascia lata graft with fibrin glue had been used for first surgery at intracranial repair in this failure patients. This failure rate was not significantly higher than that of patients in which fibrin glue was not used (P < 0.192). The use of fibrin glue to fixate fascia lata graft did not benefit the outcome.
DISCUSSION The most important complication of CSF rhinorrhea is meningitis. The CSF rhinorrhea may not be clinically apparent and manifest months or years later as meningitis.5,7 Delayed or recurrent CSF rhinorrhea almost never stops without operative treatment, and the risk for meningitis becomes high. The risk for meningitis occuring within 10 years is reduced from 85% to 7% when a dural repair is undertaken.8 This article reports our experience in surgical treating 13 patients with recurrent meningitis due to CSF rhinorrhea after cranial trauma. The CSF rhinorrhea after craniofacial trauma can ocur in many sites involving the skull base and midfacial region. The CSF enters the nasal cavity through defects in both dura mater and bone in relation to the frontal, ethmoid, or sphenoid sinuses or along the cribriform plate.1,9 Fractures of the anterior cranial base were classified into different types by the different authors. Asano et al10 classified that CT scans adjusted to bone density disclosed 3 fractures types: type 1, penetrating fractures through the orbita or the ethmoid sinus; type 2, simple or multiple linear fractures in the frontal or ethmoid sinus, the roof of the orbita, or over the planum sphenoidale through the roof of the orbita; type 3, extensive comminuted fractures of the frontal and ethmoid sinuses and 1 or both orbital roofs with or without fracture of the sella turcica. These authors found a relationship between the fracture type and the admission Glasgow Coma Scale (GCS) score and the extent of dural tears contributing to CSF rhinorrhea. In addition, large dural lacerations were always present in patients with type 3 fractures. Sakas et al11 presented a classification of trauma to the cranial base, based on observation in 48 cases for surgical dural repair and conservative treatment. Twenty-eight patients underwent surgical repair, and 20 patients were managed conservatively. There were intracranial infections (meningitis or abscess) in 15 patients. These authors classified those into the following four types: type I (cribriform), linear fracture through the cribriform plate without involvement of the ethmoid or frontal sinuses; type II (frontoethmoidal), fracture through the horizontal portion of the anterior skull base, directly related to the medial frontal and/or ethmoid sinuses, excluding the cribriform plate; type III (lateral frontal), fracture involving the lateral aspect of the frontal sinus and superomedial wall of the orbit; type IV
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
171
Yaldız et al
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
(complex), any combination of the other 3 types. Nonetheless, these authors classified the fracture into 2 subgroups: (1) small, measuring less than 1 cm; and (2) large, measuring more than 1 cm. The number of patients who developed intracranial infection in each fracture type was as follows: type I, 3 of 7; type II, 4 of 12; type III, 3 of 13; and type IV, 5 of 16. These authors indicate that large fracture, location in proximity to the midline, and CSF rhinorrhea lasting more than 8 days have high risk for infection. Large type I fractures are most prone to infection, whereas small type III fractures showed the least risk for infection. Large type I fracture with transient rhinorrhea lasting 5 to 8 days and large type II fracture with prolonged rhinorrhea lasting more than 8 days should be surgically repaired. Madhusudan et al12 proposed a clinicoradiographic classification for skull base fracture associated with midfacial injuries and differentiated 3 main types: central (type 1), lateral (type 2), and combined (type 3). Subregions were divided into frontal including cranial vault, basal with floor of the anterior cranial fossa, ethmoid and lateral orbital walls, and frontobasal with both frontal and basal regions. The fracture was called impure if there was a midfacial fracture and pure if without. Moreover, 48.7% of the patients with impure frontobasal fractures showed CSF leaks, whereas only 16.7% had it with pure frontobasal fractures. The most common fracture associated with CSF rhinorrhea was that of the frontal bone involving the frontal sinus, followed by orbital and ethmoid fractures in our patients. Three patients with dural defect larger than 1 cm were seen during surgery. The size of bone and dural defect seems to be an important prognostic factor for episodes of meningitis and brain abscess in our patients. Choi and Spann13 and Bernal-Sprekelsen et al5 showed that the conservative or surgical repair did not quarantee the prevention of meningitis. Besides, delayed CSF rhinorrhea or meningitis may develop many years later. The underdiagnosis of single or multiple CSF fistulas, the lack of adequate separation of cranial cavity, the development of posttraumatic hydrocephalus, and altered paranasal sinus drainage are associated with an increased incidence of delayed meningitis.5 Rocchi et al8 identified a relatively high risk for meningitis associated with dural fistulae. Even in those patients who were treated conservatively on a temporary basis while awaiting progression of cerebral edema or vegetative parameters (surgical treatment from 12 to 25 days after trauma), 2 of the 8 patients developed meningitis. These authors noted that the conservative therapy of CSF fistulae is only successful in selected cases. The preserve of specific (bone displacement more than 1 cm, location in proximity to the midline, involvement of the cribriform plate), evidence of encephalocele or meningocele, and persistence of rhinorrhea require surgical repair. Yilmazlar et al2 showed that, if a patient has a GCS scores of greater than 8 at admission, the prognosis is favorable, and the risk for meningitis is low. Therefore, patients with rhinorrhea who present with GCS scores of greater than 8 and have associated cranial pathologic conditions should initially be treated by surgery, and skull base defects should be repaired in the same session. Jennett and Teasdale14 reported that repair of a basal dural tear is more important in the prevention of meningitis than the treatment of CSF leaks because compound skull fractures can be complicated with meningitis without associated CSF leaks. A lumbar spinal drain may be effective in reducing the hydrostatic pressure, allowing the CSF leak to close by an inflammatory response.4 Success rates of 84% to 94% have been reported for lumbar spinal drain when used as a treatment of CSF leaks.15 It has been recommended to keep a lumbar spinal drain in place for 4 to 10 days, draining approximately 150 mL of CSF daily.4 Endoscopic repair of CSF leaks is accepted as the treatment of choice for repair of skull base defects resulting in CSF rhinorrhea. Wigand16 reported the first endoscopic repair of CSF leak in 1981. Endonasal endoscopic approach can be preferred for the clossure of
172
uncomplicated CSF rhinorrhea, located at the anterior or posterior ethmoid roof and in the spheoid sinus. During the past decade, rates of success after the first attempt at repair have ranged from 85% to 90%. Success after the second attempt at repair has ranged from 92% to 97%.17,18 A bifrontal craniotomy provides excellent exposure of the floor of the anterior cranial fossa for both intradural and extradural approaches. Transcranial approaches include the frontal craniotomy described by Dandy19 in 1926, transnasal approach with cranialization of the frontal sinus, and suprasinus transfrontal approach with lateral extension if it is necessary to achieve posterior regions.8 An intracranial approach allows direct visualization of the dural tear, inspection of any adjacent brain injury, and resection of any coexisting intracranial pathology. In addition, a free autologous graft or a pericranialgaleal flap may be necessary to seperate the dural rent from the paranasal sinuses. Intracranial approaches should be reserved for more complicated CSF rhinorrhea, which results from extensive comminuted fractures of the anterior cranial base and is accompanied with intracranial complications. Gassner et al20 reported that the most successful sealing material for the repair of CSF rhinorrhea were free grafts, for example, autogenous muscle, temporalis fascia, or fascia lata. Some authors recommended the use of free grafts without fibrin glue as an excellent option for the repair of applicable defects. In our series, fascia lata grafts fixated with fibrin glue failed more frequently than those without fibrin glue. Among the 13 patients who were not treated succesfully by the first surgery, three declined a second surgery. Fibrin glue was applied in 3 of the 13 patients. On the basis of Gassner et al's20 study as well as our own results, we currently do not recommend the use of fibrin glue during the intracranial repair of posttraumatic CSF rhinorrhea associated with recurrent meningitis. In our series, the presence of large bone defect and dural defect of more than 1 cm was the most common reason for failure to repair a CSF rhinorrhea with recurrent meningitis. There was a limitation for intracranial repair of posttraumatic CSF rhinorrhea in our study, because first repair surgery was not homogenous and lumbar spinal drain or lumboperitoneal shunt was not used at first surgery in our patients.
CONCLUSIONS Explorative craniotomy and closure of leaks or defects at the skull base should be used in these cases. There does seem to be a significant decrease in the incidence of meningitis if surgical dural repair is performed. The size of bone and dural defect seems to be an important prognostic factor of posttraumatic CSF rhinorrhea associated with recurrent meningitis. In our series, the presence of brain abscess, craniectomy defect, and large dural defect was the most common reason for failure to repair a CSF rhinorrhea with recurrent meningitis. The use of fibrin glue to fixate fascia lata graft did not improve the result in this series. Surgical outcomes in these patients could be improved by precise preoperative leakage location with optimal surgical approach and the surgeon's skill and experience. A transcranial surgical approach is required in patients with anterior cranial fossa fractures who present with recurrent meningitis.
REFERENCES 1. Goel G, Ravishankar S, Jayakumar PN, et al. Intrathecal gadolinium-enhanced magnetic resonance cisternography in cerebrospinal fluid rhinorrhea: road ahead? J Neurotrauma 2007;24:1570–1575 2. Yilmazlar S, Arslan E, Kocaeli H, et al. Cerebrospinal fluid leakage complicating skull base fractures: analysis of 81 cases. Neurosurg Rev 2006;29:64–71 3. McMains KC, Gross CW, Kountakis SE. Endoscopic management of cerebrospinal fluid rhinorrhea. Laryngoscope 2004;114:1833–1837
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
4. Brandt MT, Jenkins WS, Fattahi TT, et al. Cerebrospinal fluid: implications in oral and maxillofacial surgery. J Oral Maxillofac Surg 2002;60:1049–1056 5. Bernal-Sprekelsen M, Bleda-Vazquez C, Carrau RL. Ascending meningitis secondary to traumatic cerebrospinal fluid leaks. Am J Rhinol 2000;14:257–259 6. Zapalac JS, Marple BF, Schwade ND. Skull base cerebrospinal fluid fistulas: a comprehensive diagnostic algorithm. Otolaryngol Head Neck Surg 2002;126:669–676 7. Russell T, Cummins BH. Cerebrospinal fluid rhinorrhea thirty four years after trauma: a case report and review of literature. Neurosurgery 1984;15:705–708 8. Rocchi G, Caroli E, Belli E, et al. Severe craniofacial fractures with frontobasal involvement and cerebrospinal fluid fistula: indications for surgical repair. Surg Neurol 2005;63:559–564 9. Clemenza JW, Kaltman SI, Diamond DL. Craniofacial trauma and cerebrospinal fluid leakage: a retrospective clinical study. J Oral Maxillofac Surg 1995;53:1004–1007 10. Asano T, Ohno K, Takada Y, et al. Fractures of the floor of the anterior cranial fossa. J Trauma 1995;39:702–706 11. Sakas DE, Beale DJ, Ameen AA, et al. Compound anterior cranial base fractures: classification using computerized tomography scanning as a basis for selection of patients for dural repair. J Neurosurg 1998;88:471–477
CSF Rhinorrhea With Recurrent Meningitis
12. Madhusudan G, Sharma RK, Khandelwal N, et al. Nomenclature of frontobasal trauma: a new clinicoradiographic classification. Plast Reconstr Surg 2006;117:2382–2388 13. Choi D, Spann R. Traumatic cerebrospinal fluid leakage: risk factors and the use of prophlactic antibiotics. Br J Neurosurg 1996;10:571–575 14. Jennett B, Teasdale G. Management of head injuries: open injuries. Philadelphia: FA Davis, 1981 15. Bell RB, Dierks EJ, Homer L, et al. Management of cerebrospinal fluid leak associated with craniomaxillofacial trauma. J Oral Maxillofac Surg 2004;62:676–684 16. Wigand ME. Trannasal ethmoidectomy under endoscopical control. Rhinology 1981;19:7–15 17. Hegazy HM, Carrau RL, Snyderman CH, et al. Trannasal endoscopic repair of cerebrospinal fluid rhinorrhea: a meta-analysis. Laryngoscope 2000;110:1166–1172 18. Mirza S, Thaper A, McClelland L, et al. Sinonasal cerebrospinal fluid leaks: management of 97 patients over 10 years. Laryngoscope 2005;115:1774–1777 19. Dandy WE. Pneumocephalus (intracranial pneumatocele or aerocele). Arch Surg 1995;12:949–982 20. Gassner HG, Ponikau JU, Sherris DA, et al. CSF rhinorrhea: 95 consecutive surgical cases with long term follow up at the Mayo Clinic. Am J Rhinol 1999;13:439–447
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
173
Intracranial Repair of Posttraumatic Cerebrospinal Fluid Rhinorrhea Associated With Recurrent Meningitis Can Yaldiz, MD,* Nail Ozdemir, MD,† Onur Yaman, MD,‡ İsmail Ertan Seyin, MD,‡ and Serdar Oguzoglu, MD‡ Objective: The purposes of this study are to assess the efficacy of our intracranial surgery and evaluate the association between failure after first surgical repair and the risk factors that have been applied on a group of 13 patients affected by posttraumatic cerebrospinal fluid rhinorrhea associated with recurrent meningitis. Methods: We retrospectively collected data on 13 patients referred to our institution. All patients had history of head trauma and experienced 2 or more episodes of meningitis. Results: Three of the 13 patients had craniectomy defect due to previous trauma and surgery, 9 patients had linear fracture, and 1 patient had no apparent fracture line on preoperative radiologic evaluation. Ten of the 13 patients had identified frontal bone fracture involving the frontal sinus during surgery. Dural tear was identified intradurally and was repaired using a fascia lata graft with or without fibrin glue. Fibrin glue was applied over the suture in 7 patients. Three of the 13 patients had large dural defects. Conclusions: The size of bone and dural defect seems to be an important prognostic factor of episodes of meningitis. The use of fibrin glue to fixate fascia lata graft did not benefit the outcome. Key Words: Cerebrospinal fluid rhinorrhea, head trauma, meningitis, surgical repair (J Craniofac Surg 2015;26: 170–173)
P
osttraumatic cerebrospinal fluid (CSF) rhinorrhea is caused by tearing of the arachnoid and dura mater associated with skull fractures. Approximately 80% of all cases of CSF rhinorrhea occur after head injury, and this injury commonly occurs at the floor of the anterior cranial fossa. The CSF rhinorrhea implies leakage of CSF through the meninges and skull base into the paranasal sinuses with subsequent leakage through the anterior nasal apertures.1,2 The incidence of meningitis from CSF rhinorrhea is reported to be 10% and up to 40% during long-term follow-up.3 The mortality rate of
From the *Department of Neurosurgery, Sakarya University Training and Research Hospital, Sakarya; †Department of Neurosurgery, İzmir Tepecik Training and Research Hospital, İzmir; and ‡Department of Neurosurgery, İzmir Atatürk Training and Research Hospital, İzmir, Turkey. Received March 16, 2014. Accepted for publication June 20, 2014. Address correspondence and reprint requests to Can Yaldiz, MD, Department of Neurosurgery, Sakarya University Training and Research Hospital, Şirinevler Mh., 54100 Sakarya, Turkey; E-mail: [email protected] Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.jcraniofacialsurgery.com). The authors report no conflicts of interest. Copyright © 2014 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000001181
170
posttraumatic CSF rhinorrhea associated with meningitis is approximately 10%.2,4 Meningitis may develop acutely or even decades after the trauma.5 Because of the attendant risk for meningitis, brain abscess, and pneumocephalus, all persistent CSF rhinorrhea should be repaired.6 The purposes of this retrospective analysis are to determine the relationship between the fracture type, bone defect, dural defect, pneumocephalus, brain abscess, precise preoperative localization, fascia lata graft, fibrin glue, and persistent CSF rhinorrhea and assess the clinical outcomes and prognostic factors of patients admitted to our clinic with posttraumatic CSF rhinorrhea associated with recurrent meningitis who were managed by surgical intracranial repair.
MATERIALS AND METHODS Thirteen patients with clinical CSF rhinorrhea were included in this study at the Neurosurgery Clinic of İzmir Atatürk Research and Training Hospital from September 2003 to June 2009. All patients had cranial trauma and episodes of meningitis before our study. None of the patients had clinically detectable meningitis at admission. Nonetheless, each patient had received antibiotic therapy during episodes of meningitis. Retrospective analysis of data including demographic details, etiology, location of CSF leak, imaging results, surgical techniques, complications, recurrence, and surgical outcomes were examined. Indications for surgery included chronic occult rhinorrhea and recurrent meningitis. Iatrogenic leaks from previous neurosurgical or rhinosurgical surgery (such as removal of a pituitary adenoma, paranasal sinus surgery, endonasal procedure, orbital decompression) and spontaneous causes (such as congenital malformations or tumors) were excluded.
RESULTS Patient Data During the studied period, 13 patients were referred with apparent CSF rhinorrhea and recurrent meningitis due to chronic rhinorrhea. The patients consisted of 12 males and 1 female, ranging in age from 11 to 74 years (mean age, 28.6 y); 10 patients were younger than 30 years. History of a significant trauma to the head and/or obvious findings at surgery consistent with a traumatic origin were included in this study. Of the total 13 patients, 9 had been treated nonsurgically, and 4 had undergone cranial surgery. Surgery was necessary in 4 patients who had skull base fractures accompanied by a penetrating injury or who had depressed cranial fractures associated with cerebral pathology. Eight patients had been injured in traffic accidents, 4 patients had fallen, and 1 patient experienced a gunshot wound. A total of 13 patients had occult CSF rhinorrhea at admission, and all these individuals presented with recurrent meningitis after the original trauma. We reviewed the previous medical records of 13 patients who were treated for bacterial meningitis due to CSF rhinorrhea. In these patients, the mean interval from head injury to recurrent meningitis was 36 months (minimum, 1 mo; maximum, 132 mo). All patients experiences 2 or more episodes of bacterial
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
meningitis. The number of patients with repeated episodes of meningitis in this study was as follows: 2 attacks, 5 of 13; 3 attacks, 3 of 13; and 5 attacks, 3 of 13. Two patients in our series developed cerebral abscesses after meningitis, and all had been treated succesfully. One patient with cerebral abscess experienced a gunshot wound.
Preoperative Evaluation The mean Glasgow Outcome Scale score at admission was 4.61 in all patients. The Glasgow Outcome Scale was favorable for all 13 of the patients; none of the patients were severely disabled. A total of 6 patients had cranial nerve injuries (olfactory nerve, 2 patients; optic nerve, 4 patients). All patients underwent noncontrast high-resolution computed tomography (CT) with bone window images. Six patients required magnetic resonance imaging (MRI) to define the skull base defect. Three of the 13 patients had craniectomy defect and pneumocephalus due to previous trauma and surgery (including the 2 cerebral abscesses), and 9 patients had linear fracture. One patient had not identified fracture line or other pathology with CT and MRI (see Supplemental Digital Content, Table 1, http:// links.lww.com/SCS/A100).
Treatment A bicoronal scalp incision was made along the hairline. The flap is elevated anteriorly beyond the supraorbital ridges and glabella. Laterally, the dissection was done superficial to the temporalis fascia. The supraorbital nerves and vessels are identified and carefully dissected from the supraorbital notch. Wide opposure of the ethmoidal roof, the whole frontal region, the orbital roofs, and lateral walls up the orbit apex is performed. Sometimes, it was possible to remove frontal and orbital bone segments along fracture lines. Then, a bilateral frontal craniotomy and dural incision were made. With the aid of an operative microscope, gentle retraction of the frontal lobes was used to allow visualization of the bone and dural defects and the herniated brain tissue. Dural tear was identified intradurally and could not be closed primarily; it was repaired using a fascia lata graft under the operating microscope. Care was taken to try to obtain a water-tight closure. Fibrin glue was applied over the suture in 7 patients. Other 6 patients had used only fascia lata graft. Ten of the 13 patients had identified frontal bone fracture involving the frontal sinus. Three of the 13 patients had a dural defect of more than 1 cm. The anterior fossa was found to have numerous small 2- to 3-mm indentations, and approximately 2 cm behind the posterior wall of the frontal sinus, an outpouching of dura was found with a dural tear and herniation of brain tissue. The frontal sinuses are meticulously reconstructed. These patients were treated using mucozectomy and fat obliteration of the frontal sinuses associated with fascia lata duraplasty (see Supplemental Digital Content, Table 1, http://links.lww.com/SCS/A100). Lumbar drains were not used routinely in our patients. The mean postoperative stay in the intensive care unit was 48 hours, and the mean hospital stay was 26 days. Two patients had postoperative temporary nontension pneumocephalus. There were no meningitis, epidural or intradural hematoma, frontal lobe infarct or contusion, and seizures after surgery. The mean followup was 4.4 years, with a range of 6 months to 6 years. Success was achieved after the first surgery in 10 of the 13 patients (77%). In 3 patients who were reoperated on for recurrent rhinorrhea, a lumboperitoneal shunt was placed during the second surgery. Success was achieved in all 3 patients after a second surgery at repair (see Supplemental Digital Content, Table 1, http://links.lww. com/SCS/A100). In these cases of failure after first surgical repair, the risk factors evaluated for possible association included site and size defect, pneumocephalus, brain abscess, precise preoperative localization, and
CSF Rhinorrhea With Recurrent Meningitis
materials used for repair. Chi-squared test was used for statistical analysis. Association between failure after first surgical repair and the risk factors above was evaluated based on Fisher exact test. All calculated P values were 2-sided, and Ps < 0.05 were considered statistically significant. There were 2 cases of brain abscess after cranial trauma surgery in this reoperated group. Association between failure and presence of brain abscess was statistically significant (P < 0.038). There were 3 cases of a craniectomy defect and large dural defect (measuring more than 1 cm) in this reoperated group. Association between failure and presence of craniectomy defect was statistically significant (P < 0.038). Association between failure and large dural defect was statistically significant (P < 0.038). If the patient has had no history of craniotomy or brain abscess, rhinorrhea usually did not recur after the first operation. Intraoperative localization was confirmed by CT and/or MRI in only 5 patients. Pneumocephalus was related to an increased incidence of brain abscess. However, in patients with multiple sites of fracture, there was no significant increased number of recurrent meningitis. Fascia lata graft with fibrin glue was applied in 7 of the 13 patients. Three of the 13 patients required reoperation for recurrent leakage, and fascia lata graft with fibrin glue had been used for first surgery at intracranial repair in this failure patients. This failure rate was not significantly higher than that of patients in which fibrin glue was not used (P < 0.192). The use of fibrin glue to fixate fascia lata graft did not benefit the outcome.
DISCUSSION The most important complication of CSF rhinorrhea is meningitis. The CSF rhinorrhea may not be clinically apparent and manifest months or years later as meningitis.5,7 Delayed or recurrent CSF rhinorrhea almost never stops without operative treatment, and the risk for meningitis becomes high. The risk for meningitis occuring within 10 years is reduced from 85% to 7% when a dural repair is undertaken.8 This article reports our experience in surgical treating 13 patients with recurrent meningitis due to CSF rhinorrhea after cranial trauma. The CSF rhinorrhea after craniofacial trauma can ocur in many sites involving the skull base and midfacial region. The CSF enters the nasal cavity through defects in both dura mater and bone in relation to the frontal, ethmoid, or sphenoid sinuses or along the cribriform plate.1,9 Fractures of the anterior cranial base were classified into different types by the different authors. Asano et al10 classified that CT scans adjusted to bone density disclosed 3 fractures types: type 1, penetrating fractures through the orbita or the ethmoid sinus; type 2, simple or multiple linear fractures in the frontal or ethmoid sinus, the roof of the orbita, or over the planum sphenoidale through the roof of the orbita; type 3, extensive comminuted fractures of the frontal and ethmoid sinuses and 1 or both orbital roofs with or without fracture of the sella turcica. These authors found a relationship between the fracture type and the admission Glasgow Coma Scale (GCS) score and the extent of dural tears contributing to CSF rhinorrhea. In addition, large dural lacerations were always present in patients with type 3 fractures. Sakas et al11 presented a classification of trauma to the cranial base, based on observation in 48 cases for surgical dural repair and conservative treatment. Twenty-eight patients underwent surgical repair, and 20 patients were managed conservatively. There were intracranial infections (meningitis or abscess) in 15 patients. These authors classified those into the following four types: type I (cribriform), linear fracture through the cribriform plate without involvement of the ethmoid or frontal sinuses; type II (frontoethmoidal), fracture through the horizontal portion of the anterior skull base, directly related to the medial frontal and/or ethmoid sinuses, excluding the cribriform plate; type III (lateral frontal), fracture involving the lateral aspect of the frontal sinus and superomedial wall of the orbit; type IV
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
171
Yaldız et al
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
(complex), any combination of the other 3 types. Nonetheless, these authors classified the fracture into 2 subgroups: (1) small, measuring less than 1 cm; and (2) large, measuring more than 1 cm. The number of patients who developed intracranial infection in each fracture type was as follows: type I, 3 of 7; type II, 4 of 12; type III, 3 of 13; and type IV, 5 of 16. These authors indicate that large fracture, location in proximity to the midline, and CSF rhinorrhea lasting more than 8 days have high risk for infection. Large type I fractures are most prone to infection, whereas small type III fractures showed the least risk for infection. Large type I fracture with transient rhinorrhea lasting 5 to 8 days and large type II fracture with prolonged rhinorrhea lasting more than 8 days should be surgically repaired. Madhusudan et al12 proposed a clinicoradiographic classification for skull base fracture associated with midfacial injuries and differentiated 3 main types: central (type 1), lateral (type 2), and combined (type 3). Subregions were divided into frontal including cranial vault, basal with floor of the anterior cranial fossa, ethmoid and lateral orbital walls, and frontobasal with both frontal and basal regions. The fracture was called impure if there was a midfacial fracture and pure if without. Moreover, 48.7% of the patients with impure frontobasal fractures showed CSF leaks, whereas only 16.7% had it with pure frontobasal fractures. The most common fracture associated with CSF rhinorrhea was that of the frontal bone involving the frontal sinus, followed by orbital and ethmoid fractures in our patients. Three patients with dural defect larger than 1 cm were seen during surgery. The size of bone and dural defect seems to be an important prognostic factor for episodes of meningitis and brain abscess in our patients. Choi and Spann13 and Bernal-Sprekelsen et al5 showed that the conservative or surgical repair did not quarantee the prevention of meningitis. Besides, delayed CSF rhinorrhea or meningitis may develop many years later. The underdiagnosis of single or multiple CSF fistulas, the lack of adequate separation of cranial cavity, the development of posttraumatic hydrocephalus, and altered paranasal sinus drainage are associated with an increased incidence of delayed meningitis.5 Rocchi et al8 identified a relatively high risk for meningitis associated with dural fistulae. Even in those patients who were treated conservatively on a temporary basis while awaiting progression of cerebral edema or vegetative parameters (surgical treatment from 12 to 25 days after trauma), 2 of the 8 patients developed meningitis. These authors noted that the conservative therapy of CSF fistulae is only successful in selected cases. The preserve of specific (bone displacement more than 1 cm, location in proximity to the midline, involvement of the cribriform plate), evidence of encephalocele or meningocele, and persistence of rhinorrhea require surgical repair. Yilmazlar et al2 showed that, if a patient has a GCS scores of greater than 8 at admission, the prognosis is favorable, and the risk for meningitis is low. Therefore, patients with rhinorrhea who present with GCS scores of greater than 8 and have associated cranial pathologic conditions should initially be treated by surgery, and skull base defects should be repaired in the same session. Jennett and Teasdale14 reported that repair of a basal dural tear is more important in the prevention of meningitis than the treatment of CSF leaks because compound skull fractures can be complicated with meningitis without associated CSF leaks. A lumbar spinal drain may be effective in reducing the hydrostatic pressure, allowing the CSF leak to close by an inflammatory response.4 Success rates of 84% to 94% have been reported for lumbar spinal drain when used as a treatment of CSF leaks.15 It has been recommended to keep a lumbar spinal drain in place for 4 to 10 days, draining approximately 150 mL of CSF daily.4 Endoscopic repair of CSF leaks is accepted as the treatment of choice for repair of skull base defects resulting in CSF rhinorrhea. Wigand16 reported the first endoscopic repair of CSF leak in 1981. Endonasal endoscopic approach can be preferred for the clossure of
172
uncomplicated CSF rhinorrhea, located at the anterior or posterior ethmoid roof and in the spheoid sinus. During the past decade, rates of success after the first attempt at repair have ranged from 85% to 90%. Success after the second attempt at repair has ranged from 92% to 97%.17,18 A bifrontal craniotomy provides excellent exposure of the floor of the anterior cranial fossa for both intradural and extradural approaches. Transcranial approaches include the frontal craniotomy described by Dandy19 in 1926, transnasal approach with cranialization of the frontal sinus, and suprasinus transfrontal approach with lateral extension if it is necessary to achieve posterior regions.8 An intracranial approach allows direct visualization of the dural tear, inspection of any adjacent brain injury, and resection of any coexisting intracranial pathology. In addition, a free autologous graft or a pericranialgaleal flap may be necessary to seperate the dural rent from the paranasal sinuses. Intracranial approaches should be reserved for more complicated CSF rhinorrhea, which results from extensive comminuted fractures of the anterior cranial base and is accompanied with intracranial complications. Gassner et al20 reported that the most successful sealing material for the repair of CSF rhinorrhea were free grafts, for example, autogenous muscle, temporalis fascia, or fascia lata. Some authors recommended the use of free grafts without fibrin glue as an excellent option for the repair of applicable defects. In our series, fascia lata grafts fixated with fibrin glue failed more frequently than those without fibrin glue. Among the 13 patients who were not treated succesfully by the first surgery, three declined a second surgery. Fibrin glue was applied in 3 of the 13 patients. On the basis of Gassner et al's20 study as well as our own results, we currently do not recommend the use of fibrin glue during the intracranial repair of posttraumatic CSF rhinorrhea associated with recurrent meningitis. In our series, the presence of large bone defect and dural defect of more than 1 cm was the most common reason for failure to repair a CSF rhinorrhea with recurrent meningitis. There was a limitation for intracranial repair of posttraumatic CSF rhinorrhea in our study, because first repair surgery was not homogenous and lumbar spinal drain or lumboperitoneal shunt was not used at first surgery in our patients.
CONCLUSIONS Explorative craniotomy and closure of leaks or defects at the skull base should be used in these cases. There does seem to be a significant decrease in the incidence of meningitis if surgical dural repair is performed. The size of bone and dural defect seems to be an important prognostic factor of posttraumatic CSF rhinorrhea associated with recurrent meningitis. In our series, the presence of brain abscess, craniectomy defect, and large dural defect was the most common reason for failure to repair a CSF rhinorrhea with recurrent meningitis. The use of fibrin glue to fixate fascia lata graft did not improve the result in this series. Surgical outcomes in these patients could be improved by precise preoperative leakage location with optimal surgical approach and the surgeon's skill and experience. A transcranial surgical approach is required in patients with anterior cranial fossa fractures who present with recurrent meningitis.
REFERENCES 1. Goel G, Ravishankar S, Jayakumar PN, et al. Intrathecal gadolinium-enhanced magnetic resonance cisternography in cerebrospinal fluid rhinorrhea: road ahead? J Neurotrauma 2007;24:1570–1575 2. Yilmazlar S, Arslan E, Kocaeli H, et al. Cerebrospinal fluid leakage complicating skull base fractures: analysis of 81 cases. Neurosurg Rev 2006;29:64–71 3. McMains KC, Gross CW, Kountakis SE. Endoscopic management of cerebrospinal fluid rhinorrhea. Laryngoscope 2004;114:1833–1837
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 26, Number 1, January 2015
4. Brandt MT, Jenkins WS, Fattahi TT, et al. Cerebrospinal fluid: implications in oral and maxillofacial surgery. J Oral Maxillofac Surg 2002;60:1049–1056 5. Bernal-Sprekelsen M, Bleda-Vazquez C, Carrau RL. Ascending meningitis secondary to traumatic cerebrospinal fluid leaks. Am J Rhinol 2000;14:257–259 6. Zapalac JS, Marple BF, Schwade ND. Skull base cerebrospinal fluid fistulas: a comprehensive diagnostic algorithm. Otolaryngol Head Neck Surg 2002;126:669–676 7. Russell T, Cummins BH. Cerebrospinal fluid rhinorrhea thirty four years after trauma: a case report and review of literature. Neurosurgery 1984;15:705–708 8. Rocchi G, Caroli E, Belli E, et al. Severe craniofacial fractures with frontobasal involvement and cerebrospinal fluid fistula: indications for surgical repair. Surg Neurol 2005;63:559–564 9. Clemenza JW, Kaltman SI, Diamond DL. Craniofacial trauma and cerebrospinal fluid leakage: a retrospective clinical study. J Oral Maxillofac Surg 1995;53:1004–1007 10. Asano T, Ohno K, Takada Y, et al. Fractures of the floor of the anterior cranial fossa. J Trauma 1995;39:702–706 11. Sakas DE, Beale DJ, Ameen AA, et al. Compound anterior cranial base fractures: classification using computerized tomography scanning as a basis for selection of patients for dural repair. J Neurosurg 1998;88:471–477
CSF Rhinorrhea With Recurrent Meningitis
12. Madhusudan G, Sharma RK, Khandelwal N, et al. Nomenclature of frontobasal trauma: a new clinicoradiographic classification. Plast Reconstr Surg 2006;117:2382–2388 13. Choi D, Spann R. Traumatic cerebrospinal fluid leakage: risk factors and the use of prophlactic antibiotics. Br J Neurosurg 1996;10:571–575 14. Jennett B, Teasdale G. Management of head injuries: open injuries. Philadelphia: FA Davis, 1981 15. Bell RB, Dierks EJ, Homer L, et al. Management of cerebrospinal fluid leak associated with craniomaxillofacial trauma. J Oral Maxillofac Surg 2004;62:676–684 16. Wigand ME. Trannasal ethmoidectomy under endoscopical control. Rhinology 1981;19:7–15 17. Hegazy HM, Carrau RL, Snyderman CH, et al. Trannasal endoscopic repair of cerebrospinal fluid rhinorrhea: a meta-analysis. Laryngoscope 2000;110:1166–1172 18. Mirza S, Thaper A, McClelland L, et al. Sinonasal cerebrospinal fluid leaks: management of 97 patients over 10 years. Laryngoscope 2005;115:1774–1777 19. Dandy WE. Pneumocephalus (intracranial pneumatocele or aerocele). Arch Surg 1995;12:949–982 20. Gassner HG, Ponikau JU, Sherris DA, et al. CSF rhinorrhea: 95 consecutive surgical cases with long term follow up at the Mayo Clinic. Am J Rhinol 1999;13:439–447
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
173
