Portuguese Version

Year:  2002  Vol. 68   Ed. 6 - (2º)

Artigos Originais

Pages: 788 to 792

Localization of skull base cerebrospinal fluid leaks by using intrathecal fluorescein in a hypodense solution during the surgery

Author(s): Roberto E. S. Guimarães 1,
Helena M. G. Becker 1,
Celso G. Becker 2,
Paulo F. B. T. Crosara 3,
Denise U. Gonçalves 4,
Alcino L. Silva 5

Keywords: cerebrospinal rinorrhea, cerebrospinal fluid injection, diagnostic fluorescein

Abstract:
The utilization of intrathecal fluorescein in the surgery, when it can be identified allows the precise localization of skull base cerebrospinal fluid leaks, thus making the treatment easier. Aim: To evaluate the use of intrathecal fluorescein in hipodense solution, in the treatment of CSF rinorrhea. Study design: Clinical retrospective. Material and Method: In this study, 23 patients with CSF rinorrhea were operated by the author from October 1997 to May 1999 using intrathecal fluorescein in a hypodense solution during the surgery. In relation to gender and age, 10 (43,5%) of the patients were female, 13 (56,5%) male, varying from 4 to 57 years old. All the fistulae were unilateral, being 15 (65%) of them traumatic and 8 (35%) originated from nontraumatic causes. The hypodense solution was obtained from the dilution of 0,5 cm3 of fluorescein into 10 cm3 of distilled water. The density of the obtained solution is in order of 1001. Always less dense than the CSF fluid which density is in the order of 1004 to 1006. Results: This hypodense solution has prover to be safe in its use, since no adverse reaction has been reported, as well as there were no chemical or cytological alterations in any of the studied cases. Conclusions: This method has the advantage of avoiding the Trendelenburg position for approximately 40 minutes during the surgery, it makes easier the use of fluorescein in patients with CSF hypotension and, eventually, tums an inactive fistulae into an active ore, which is desirable during the surgery, since it allows the localization of the fistulae and the verification of it's effective treatment.

INTRODUCTION

Cerebrospinal fluid rhinorrhea (CSF rhinorrhea), classically defined as the presence of cerebrospinal fluid (CSF) in the nasal cavity, was first described and related to cranial trauma by surgeons in the Middle Ages. Thompson (1899)1 described that in addition to traumatic CSF rhinorrhea, there was a subgroup of spontaneous CSF rhinorrhea cases of unknown origin. From that time, many classifications were proposed, and Ommaya et al. (1968)2 made a classification that was very comprehensive and accepted by many authors, dividing CSF rhinorrhea into traumatic and non-traumatic and subdividing the latter into non-traumatic with normal CSF pressure and non-traumatic with CSF hypertension.

In the presence of clear nasal rhinorrhea, usually unilateral, there are two questions to be answered:

• Is it CSF rhinorrhea? - etiological diagnosis.
• If the answer is yes, where is it? - topographic diagnosis.

These two questions should always be answered, even though it is extremely difficult to have answers in some specific cases.

The history of the patients is the most important fact in etiological diagnosis of CSF rhinorrhea. The history of trauma within the previous months or years, be it surgical or accidental, should be seriously considered (McCoy, 1963)3, as well as the presence of seizures or meningitis, because in such conditions, in the event of pneumoccocal meningitis, the patients should be seen as carrier of CSF rhinorrhea, even if it is not evident (Crawford, 19944; Koso-Thomas & Harley, 19955).

Chemical analyses of the nasal liquid are important for the definition of CSF rhinorrhea. One of the main methods is quantitative dosage of glucose. A concentration of 30mg/dl or greater of glucose in the collected liquid confirms the presence of CSF rhinorrhea in patients with normal glucose levels (Calcaterra, 19806).

Quist-Hanssen (1961)7 recommended the test with glucose dipstick, but Gadeholt (1964)8 stated that the dipstick has a limited value because it presents 75% of false-positive reactions. Calcaterra (1980)6 emphasized that the dipstick can result in positive reaction with a small amount of glucose - 5mg/dl, which can lead to CSF leak being mistaken for lachrymal secretion or nasal discharge.

The analysis of b2 transferrin is very important for the diagnosis of CSF rhinorrhea because it is a highly specific protein presented in the human body only in CSF and perilymph, in addition to requiring a small sample to be tested (Skedros et al., 19939).

The topographic diagnosis of CSF rhinorrhea should normally be preoperative, but it is not always possible to have it, despite the great technological development of imaging techniques.

Johnson et al. (1996)10 admitted that the ideal radiological technique for the accurate location of preoperative CSF fistulae is controversial. One of the main radiological diagnostic difficulties is found in the subgroup of patients that present intermittent or inactive CSF rhinorrhea.

Computed tomography (CT) is an exam routinely conducted in patients with CSF fistulae. McCormack et al. (1990)11 and Lantz et al. (1980)12 reported that CT scan was useful in the diagnosis of CSF fistula in 50% and 53% of the patients, respectively.

Drayer et al. (1977)13 were the first ones to employ cisternography to locate CSF fistula, using metrizamide, a non-ionic water soluble product as intrathecal contrast. From then on, cisternography has been widely used as the preferred method for topodiagnosis of CSF fistula, and the results varied according to the authors and whether it was an active or inactive fistula (Prere et al., 198614).

Magnetic resonance imaging (MRI) is an important method for the diagnosis of CSF fistula in case of suspicion of cerebral abnormalities, but it is criticized because it has low spatial resolution in bone tissue (Murata et al., 199515). Toquet et al. (1998)16 advocated the use of cisternography MRI, T2 weighted MRI without contrast infection, stating that it would replace the other invasive exams.

Dying substance injected via intrathecal access and collected at the nose has been used with variable degrees of success both for diagnosis of Cogan’s syndrome and for pre and trans-operative topographic diagnosis.

Fluorescein, based on the technique proposed by Kirchner & Proud (1960)17 – injection of 1 cm3 of fluorescein diluted in 5% in 10cm3 of CSF, via intrathecal - is widely used by different surgeons for the preoperative identification of CSF fistula location and the exact trans-operative site.

Montgomery (1966)18 reported that the most important method for identification and location of CSF rhinorrhea, even paradoxical, that is, when drainage is through the auditory tube, is the intrathecal fluorescein injection, to be detected at nose or ear level. Calcaterra (1980)2 stated that in his experience, it was proven that intrathecal fluorescein was the most accurate method to locate the CSF fistula intra-operatively. The use of nasal endoscopy with intrathecal fluorescein to investigate CSF rhinorrhea was described for the first time by Messenklinger (1972)19. Since in some patients with low tension of CSF it is impossible to remove 10cm3 of the liquid, fluorescein can be diluted in 10cm3 of Hartman solution and injected via intrathecal access (Montgomery, 1973)20. The use of ultraviolet light to detect fluorescein has also been advocated, although it has been questioned by Briant & Bird (1982)21 and Calcaterra (1985)22, who reported that it is not necessary to use it to locate the CSF fistula trans-operatively. We should be extra careful to guarantee that the fluorescein used is injectable and never the topical preparation for ophthalmic use (Mattox & Kennedy, 1990)23. Upon reviewing the literature, we identified three reports of complications when using intrathecal fluorescein (Mahaley Jr. & Odom, 196624; Wallace et al., 197225 and Moseley et al., 197826). All complications were reversible and had no signs of residual deficit. The reversion time varied from 10 minutes to one week.

The widely disseminated use of intrathecal fluorescein enables the observation of the exact drainage spot of CSF. The purpose of the present study was to use trans-operative intrathecal fluorescein in hypodense solution to locate and to surgically treat CSF fistula of anterior skull base, observing how quickly it reaches the cerebral cisterns, dispensing the patient from standing in Trandelenburg position and assessing the absence of side effects to its use.

MATERIAL AND METHOD

From October 1997 to May 1999, we conducted 23 surgeries to correct CSF fistulae. Ten patients were female and 13 were male subjects, ages ranging from 4 to 57 years. All patients presented unilateral fistula, being 15 (65%) of traumatic etiology and eight (35%) of non-traumatic causes. In all operations, we used intrathecal fluorescein in hypodense solution. This solution is obtained by diluting 0.5 cm3 of fluorescein at 5% in 10 cm3 of distilled water. We created the present model to demonstrate the efficiency of intrathecal fluorescein use in hypodense solution, as well as to demonstrate quickness of action into the cerebral cisterns, when the patients were placed in reclining position,. We used two test tubes with 1.5cm in diameter and 75cm in length, which presented an orifice located 12cm from the bottom, through which we injected the liquid. We filled both tubes with saline solution that presented 1004 density, similarly to CSF, and closed them with a rubber cap through which we put the needle to enable decompression of liquid when the fluorescein injection was made. We placed both tubes simulating the surgical position, by elevating approximately 30 degrees the cap from the injection point. We injected in tube 1 the hypodense solution: 0.5cm3 of fluorescein at 5% diluted in 10cm3 of distilled water, solution with density of 1001. In tube 2, we injected the hyperdense solution: 0.5cm3 fluorescein at 5% diluted in 10cm3 of saline solution, presenting solution density of 1005. We observed the movement of liquids and noticed the great increase of fluorescein in tube 1, in which we had injected the hypodense solution, which was quick and reached the maximum point of the tube in 80 seconds. However, in tube 2, fluorescein remained practically at the same level as initially injected, with tendency to move to the bottom of the tube.

The fluorescein used for the intrathecal injection was injection sodium fluorescein at 20%, formulated by the Laboratory of Pharmacy, Medical School of UFMG, which was previously diluted for a 5% solution.
The punch was conducted before the anesthesia, except in children and more anxious elderly patients, between L3-L4 or L4-L5 with a raquianesthesia needle 15G 36”. Whenever possible, we conducted a measure of CSF pressure and removed 5cm3 of CSF for lab analysis. Next, we introduced an epidural catheter 19G into approximately 15cm, fixated on the skin. It remained there for 48 hours, when CSF was removed to decrease CSF pressure every 24 hours, as well as for chemical analysis. At the end of this procedure, we slowly injected through the epidural catheter the hypodense solution of fluorescein.
Patients were submitted to general anesthesia with orotracheal intubation and immediately placed in surgical position with the head and trunk elevated approximately 30 degrees. We started the surgery using microscopy and 0o, 30o and 45o endoscopes, helping the visualization of fluorescein and the exact location of the fistula drainage.

RESULTS

It was possible to visualize fluorescein trans-operatively in 20 (86%) of the patients, and it was not possible in three (13%) other cases.

The color of CSF postoperatively in the 23 patients was transparent after 24 hours in 2 patients (8.7%), after 48 hours in 15 (65%), and in 6 (26.1%) it remained yellowish even after 48 hours, when the catheter was removed.

The chemical and cytological CSF analysis performed in 23 patients did not show significant abnormalities when compared to results of fluorescein pre-injection and after 24 hours and 48 hours of use.

We did not detect undesirable side effects in any of the 23 operated patients after the use of intrathecal fluorescein in hypodense solution.

DISCUSSION

All the 23 patients of our series complained of unilateral rhinorrhea, being that one of them at the time of the exam did not present CSF rhinorrhea, since the CSF fistula was inactive at that time.
The report of meningitis is frequent and was detected in 27% of the patients in our series, similarly to what was observed by Lewin (1954)27 who reported occurrence of meningitis in 25% of the patients with CSF leak.

Whenever possible, the diagnosis of CSF rhinorrhea was confirmed by the dose of glucose in the nasal liquid. All the results found were similar or greater than 30mg/dl. In patients without nasal rhinorrhea, the diagnosis was clinical.

This series of 23 patients presents a gender distribution of 43.5% for female and 56.5% for male subjects, confirming a slight predominance of males. These data are similar to those reported by Park et al. (1983)28, who in a series of 42 patients with CSF leak, had 52% of male patients and 48% of females. Even though these data seem to be in accordance with the literature, they do not seem to be as significant as the predominance of females among patients with non-traumatic fistulae, in which we noticed 60% of female and 40% of male patients. This observation was also reported by Stamm (1994)29. The women found in our sample normally presented the common characteristics of short height and tendency to obesity.

Even though fluorescein has not been visualized in 3 cases (13%), two of them presented CSF fistula of traumatic etiology and had had episodes of meningitis preoperatively. They did not present active fistula in the operative period. We conducted an endonasal ethmoidectomy in one patient and external frontal sinusotomy in another patient. Surgical exploration evidenced fibrous tissue at the probable fistulae region, according to the CT scan and the preoperative cisternography. We conducted surgical reinforcement in these regions. The third patient had non-traumatic fistula with history of meningitis and during the active period of the fistula, it presented glucose level of 39mg/dl. Preoperative tests, computed tomography and cisternography were suggestive of left sphenoid fistula. We conducted left ethmoid and sphenoid exploration. Since we did not see fluorescein, we preferred to obliterate the sphenoid sinus.

In the 23 patients in whom we conducted chemical and cytological analyses of CSF fistula, before fluorescein injection and within 24 hours and 48 hours, we did not detect any significant abnormality. No patients presented any clinical abnormality resultant from the use of fluorescein. The complications resultant from the use of intrathecal fluorescein (Mahaley Jr. & Odom, 196624; Wallace et al., 197225 and Moseley et al., 197826) were reversible, without residual deficit, with recovery within 10 minutes to one week.

In conclusion, we noticed that the use of fluorescein and the technique described in this study did not present complications. It proved to be important for trans-operative diagnosis when visualized to enable precise location of the fistula; to instruct and ensure efficacy of closure; to transform an inactive fistula into an active one by increasing CSF pressure, placing immediately the patient in the surgical position and reducing the duration of operation.

Based on these facts, we believe that the technique presented in this study should be considered in the surgical treatment of CSF fistula.

REFERENCES

1. Thompson St C. The cerebro-spinal fluid: its spontaneous escape from the nose. London: Cassell & Co.; 1899. p.1- 140.
2. Ommaya AK, Di Chiro G, Baldwin M, Pennybacker JB. Non traumatic cerebrospinal fluid rhinorrhoea. J Neurol Neurosurg Psychiatry 1968;31:214-225.
3. McCoy G. Cerebrospinal rhinorrhea: a comprehensive review and a definition of the responsibility of the rhinologist in diagnosis and treatment. Laryngoscope 1963;9:1125-1157.
4. Crawford C, Kennedy N, Weir WRC. Cerebrospinal fluid rhinorrhoea and Haemophilus influenzae meningitis 37 years after a head injury. J Infect 1994;28:93-97.
5. Koso-Thomas AK, Harley EH. Traumatic cerebrospinal fluid fistula presenting as recurrent miningitis. Otolaryngol Head Neck Surg 1995; 112(3):469-472.
6. Calcaterra TC. Extracranial surgical repair of cerebrospinal rhinorrea. Ann Otol Rhinol Laryngol 1980; 89:108-116.
7. Quist-Hanssen S. Fistula of dura in fractures involving paranasal sinuses, complications and treatment. Acta Chir Scand 1961; 122:49-56.
8. Gadeholt H. The reaction of glucose-oxidase test paper in normal nasal secretion. Acta Otolaryngol 1964; 55:271-275.
9. Skedros DG, Cass SP, Hirsch BE, Kelly RH. Sources of error in use of beta-2 transferin analysis for diagnosyng perilymphatic and cerebral spinal fluid leaks. Otolaryngol. Head Neck Surg 1993; 109:861-864.
10. Johnson DB, Brennan P, Toland J, O’Dwyer AJ. Magnetic resonance imaging in the evaluation of cerebrospinal fluid fistalae. Clin Radiol 1996; 51(12):837-841.
11. Mccormack B, Cooper PR, Persky M, Rothstein S. Extracranial repair of cerebrospinal fluid fistulas: techinique and results in 37 patients. J Neurosurg 1990; 27:412-417.
12. Lantz EJ, Forbes GS, Brown ML, Laws Jr. ER. Radiology of cerebrospinal fluid rhinorrhea. AJR – Am J Roentgenol 1980; 135(5):1023-1030.
13. Drayer BP, Wilkins RH, Boehnke M, Horton JA, Rosenbaum AE. Cerebrospinal fluid rhinorrhea demonstrated by metrizamide CT cisternography. Am J Roentgenol 1977; 129:149-151.
14. Prere J, Puech JL, Deroover N, Arrue P, Tremoulet M, Manelfe C. Rhinorrhea and meningitis due to pos-traumatic osteo-meningeal defects in the anterior cranial fossa; diagnosis with water–soluble CT cisternography. J Neuroradiol 1985; 13(4):278-285.
15. Murata Y, Yamada I, Isotani E, Suzuki S. MRI in spontaneous cerebrospinal fluid rhinorrhoea. Neuroradiology 1995; 37(6):453-455.
16. Toquet J, Bordure P, Herman P, Gayet-Delacroix M, Delaroche O, Legent F. Intéret de la recherche de Beta2 transferrine et de la MR cisternographie pour le diagnostic des fistules de liquide céphalorachidien. Ann Otolaryngol Chir Cervicofac 1998; 115(5):293-298.
17. Kirchner FR, Proud GO. Method for identification and localization of cerebrospinal fluid rhinorrhea and otorrhea. Laryngoscope 1960; 70:921-931.
18. Montgomery WW. Surgery for cerebrospinal fluid rhinorrhea and otorrhea. Arch Otolaryngol 1966; 84:92-104.
19. Messerklinger W. Nasenendoskopie: Nachweis, Lokalisation und Differentialdiagnose der nasalen Liquorrehoe HNO 1972; 20:268-270.
20. Montgomery WW. Cerebrospinal fluid rhinorrhea. Otolaryngol Clin North Am 1973; 6:757-771.
21. Briant TD, Bird R. Extracranial repair of cerebrospinal fluid fistulae. J Otolaryngol 1982;11:191-197.
22. Calcaterra TC. Diagnosis and management of ethmoid cerebrospinal rhinorrea. Otolaryngol Clin North Am 1985; 18:99-105.
23. Mattox DE, Kennedy DW. Endoscopic management of cerebrospinal fluid leaks and cephaloceles. Laryngoscope 1990; 100(8):857-862.
24. Mahaley Jr. MS, Odom GL. Complication following intrathecal injection of fluorescein. J Neurosurg 1966; 25(3):298-299.
25. Wallace JD, Weintriaub MI, Mattson RH, Rosnagle R. Status epilepcticus as a complications of intrathecal fluorescein. J Neurosurg 1972; 36:659-660.
26. Moseley JI, Carton CA, Stern WE. Spectrum of complications in the use of intrathecal fluorescein. J Neurosurg 1978; 48(5):765-767.
27. Lewin W. Cerebrospinal fluid rhinorrhea in closed head injuries. Br J Surg 1954; 42:1-18.
28. Park JI, Strelzow VV, Friedman WH. Current management of cerebrospinal fluid rhinorrhea. Laryngoscope 1983; 93(10):1294-1300.
29. Stamm AC. Microcirurgia transnasal no tratamento da rinoliquorréia da região etmóidal e esfenoidal; Escola Paulista de Medicina, São Paulo; 1994 (Tese de Doutorado em Neurocirurgia).
30. Özgen T, Tekkök IH, Cila A, Erzen C. CT cisternography in evaluation of cerebrospinal fluid rhinorrhea. Neuroradiology 1990; 32(6):481-484.

---

1 Joint Professor of Otorhinolaryngology, Medical School, UFMG.
2 Assistant Professor of Otorhinolaryngology, Medical School, UFMG.
3 Doctorate studies under course, Medical School, UFMG – Volunteer preceptor of Otorhinolaryngology, Hospital das Clínicas/UFMG.
4 Ph.D., Substitute Professor in Otorhinolaryngology, Medical School, FM/UFMG.
5 Faculty Professor of Surgery, Medical School, UFMG.
Medical School, Federal University of Minas Gerais - UFMG
Address correspondence to: Roberto E. S. Guimarães – Av. Pasteur 88 4º andar
30.150-290 Belo Horizonte MG
Tel/Fax (55 31) 3222-2891 – E-mail: rguimaraes@alcance.com.br
Article submitted on June 01, 2001. Article accepted on September 19, 2002

Print: