Cerebrospinal fluid fistulae in a canine model

Cerebrospinal fluid fistulae in a canine model CHARLES A. SYMS, III, MD, MAJ, MC, USAF,MARK J. SYMS, MD, TERRENCEP. MURPHY,MD, and SAMUEL O. MASSEY,MD, COL, MC, USAF,Lackland Air Force Base, Texas, Tripler Army Medical Center, Tripler, Hawaii, and Atlanta, Georgia Various diagnostic techniques currently are used to detect the presence of a cerebrospinal fluid fistula. High resolution computerized tomography scanning with the instillation of an intrathecal nonionic contrast medium yields the most accurate diagnostic results. Occasionally, even with optimal conditions, little information is gained other than the confirmation of the presence of a fistula. Intrathecal fluorescein can provide accurate information on the exact location of the fistula. The current study was designed to refine the clinical examination for cerebrospinal fluid fistulae with the use of intrathecal fluorescein. The canine model that we used also served as a vehicle to investigate the histopathologic effects of fluorescein on the central nervous system. We modified a commonly used xenon light source to enable examination with 490 nm light. In the canine model, this allowed accurate visualization of surgically created fistulae using very low doses of intrathecal fluorescein. An examination of the histopatholagic features of the central nervous system of the canine model after acute instillation of a higher dose of fluorescein revealed microscopic changes consistent with the introduction of an irritant material. The changes induced by the chemical trauma may explain the serious neurologic sequelae sometimes seen in patients after the instillation of intrathecal fluorescein. Recommendations on the correct dosage of intrathecal fluorescein for diagnosis of cerebrospinal fluid fistulae are proposed. (Otolaryngol Head Neck Surg 1997;117:542-6).

C e r e b r o s p i n a l fluid (CSF) rhinorrhea was first described by Willis in 1676,1 and since that time, localization and repair has remained a difficult task. The efficacy of fluorescein in the clinical localization of the site of CSF leakage has been previously described. 2,3 Intravenous fluorescein has been used in research and clinically for many years with minimal side effects and infrequent allergic reactions. Apprehension surrounds the use of intrathecal fluorescein, however, because several episodes of temporary neurologic sequelae have been reported. Case reports have documented status epilepticus, aseptic meningitis, hypalgesia, numbness, weakness, opisthotonos, and cranial nerve deficits. 4-6

From the Wilford Hall U.S. Air Force Medical Center, Lackland Air Force Base (Drs. C. Syms and Massey); and the Otolaryngology-Head and Neck Surgery Service, Department of Surgery, Tripler Army Medical Center (Dr. M. Syms). Dr. Murphy is in private practice in Atlanta, Georgia. The views expressed in this article are those of the authors and do not reflect the official policy or position of the U.S. Air Force, U.S. Army, Department of Defense, or the U.S. government. Presented at the Annual Meeting of the American Academy of Otolaryngology-Headand Neck Surgery, San Diego, Calif., Sept. 18-21, 1994. Reprint requests: Mark J. Syms, MD, Otolaryngology-Head and Neck Surgery Service, Department of Surgery, Tripler Army Medical Center, Tripler, HI 96859-5000. 23/1/77576

542

No complications from the injection of intrathecal fluorescein were permanent and the reactions were thought to be idiosyncratic. 5 The original technique reported by Kirchner and Proud 3 called for 1 ml of 5% fluorescein to be injected into the lumbar intrathecal space. A year later, Kirchner 7 reported nine cases of CSF fistulae being identifed using 0.25 ml of 5% fluorescein. Once complications of intrathecal fluorescein were reported, the technique was modified by others to 0.5 ml of 5% fluorescein mixed in 5 to 10 ml of withdrawn CSF, 4,5 not 0.25 ml of 5% fluorescein as previously suggested by Kirchner. 7 The current study was designed to refine the use of intrathecal fluorescein. We investigated the pathophysiologic effect of various dosages of fluorescein in the canine model. The minimal amount of fluorescein needed to detect a CSF fistula was investigated. Finally, the efficacy of a modified light source to detect fluorescein was explored.

METHODS AND MATERIAL Sixteen healthy adults dogs of approximately the same size comprised the study group. Five of the dogs were purebred beagles and 11 were mongrels. The protocol and guidelines for this study were approved by the Wilford Hall U S A F Medical Center institutional

OtolaryngologyHead and Neck Surgery

SYMS et al. 543

Volume 117 Number 5

Fig. 1. A xenon light source (Cat #601C, Karl Storz Company, Culver City, Calif.) was modified with a 490 nm interference filter (arrows) (Ditric Optics, Hudson, Mass.), p l a c e d in the light path just proximal to the entrance of light into the fiberoptic cabte. 9

Animal Use and Care Committee. All animals were used and cared for in compliance with the Department of Defense directive 3216.1, Air Force Regulation 1692 (The Use of Animals in DOD Programs) and the National Institute of Health Publication 85-23 (Guide for the Care and Use of Laboratory Anima!,s). The animals underwent general endotracheal anesthesia, after which a cerebrospinal fistula was created. A temporal or parietal craniotomy was used to gain access to the subarachnoid space for creation of a surgical CSF leak. A CSF fistula was created by piercing the dura with a 27-gauge needle. The creation of a CSF fistula was verified by visualization of clear fluid through the surgically created dural defect. Fluorescein sodium (Fluorescite Injection; Alcon Laboratories, Fort Worth, Tex.) was injected into the intrathecal space in varying concentrations (Table 1). An aliquot of up to 10 ml of CSF was withdrawn from a lumbar spinal tap. If less than 10 ml of CSF was o6tained via the spinal tap, Elliott's B solution was added to the aliquot to raise the total volume after the addition of fluorescein to 10 ml. The diluted fluorescein was added to the 10 ml sample and injected into the intrathecal space. The fluorescein was initially injected in 0.5 ml aliquots and repeatedly diluted to lower concentrations by withdrawing additional CSF using bar-

Table 1. A m o u n t o f f l u o r e s c e i n i n j e c t e d into t h e intrathecal space and calculated concentration o f f i u o r e s c e i n in CSF Dog no.

1

2 3 4 5 6 7 8 9 10 11 12

13 14

15 16

Fluorescein (mg)

Concentration of fluorescein (mg/mO*

0.25 0.25 0.25 0.25 0.25 0.25 0.20 0,10 0.10

8.3 8.3 8.3 8.3 8.3 8.3 5.0 3.3 3.3

0,10 0.02 0.01 0.001 0,001 0.001 0.000001

x x x x x x x x x 3.3 x

10 ,3 10 -3 10 .3 10 -3 10 .3 10 -3 10 .3 10 .3 10 .3 10 -3

6.6 3.3 3.3 3,3 3.3 3.3

x 10-4 10-5 10-5 10-5 10-8

x 10 .4

x x x x

*Based on a CSF volume of 30 ml in each dog, 8

botages. 9 This consisted of repeated alternate injection and withdrawal of CSF over a period of 5 minutes and consisting of at least 30 injections until the entire 10 ml was injected. An examination of the CSF fistula was performed

OtolaryngologyHead and Neck Surgery November 1997

544 SYMS et al.

Fig. 2. Surgical wound. A temporal or parietal craniotomy was used to gain access to the subarachnoid space for the creation of the CSF fistula. The surgical CSF leak was created with a 27-gauge needle (arrowheads). The size of the fistula relative to the needle size is d e m o n strated. The surgically created fistula (arrow) is visualized using intrathecal fluorescein. Dog 16 is illustrated.

with Hopkins rod rigid telescopes (Karl Storz Company, Culver City Calif.). A xenon light source (Cat #601C, Karl Storz Company) was modified as described by Syms 1° with a 490 nm interference filter (Ditric Optics, Hudson, Mass.), placed in the light path just proximal to the entrance of light into the fiberoptic cable (Fig. 1). The CSF fistula, examined with the modified light source, was easily detected (Fig. 2). The animals were allowed to recover from general anesthesia and were monitored for the appearance of ventilator dependence, paralysis, or seizures: If a dog experienced any of these neurologic sequelae, it was immediately killed. If the animal experienced no noticeable neurologic sequelae, it was killed within 4 hours of the initial injection. Necropsy was performed on all animals. Tissues of the brain, meninges, and spinal cord were prepared for analysis by fixation in formalin, embedded in paraffin, and sectioned and stained with hematoxylin

and eosin. They were analyzed microscopically by a research pathologist. RESULTS

Cerebrospinal fluid fistulae were identified at the fistula sites of all 16 dogs using the modified light source (Fig. 1). The concentration of fluorescein ranged from 8.3 x 10-3 mg/ml to 3.3 x 10-8 mg/ml (Table 1). The fistulae were easily observed at the lowest concentration of fluorescein (Fig. 2) with the modified light source. Neurologic sequelae were observed in dogs 1 through 7 and these dogs were immediately killed. The sequelae ranged from seizures in dog 1 to ventilator dependency in dog 7. The histopathologic findings of the necropsies revealed a multifocal, diffuse, mild to severe meningoencephalomyelitis. The lesions were principally characterized by intense neutrophilic meningitis and marked perivasculitis (Fig. 3). The lesions were most severe closest to the injection site.

OtolaryngologyHead and Neck Surgery Volume 117 N u m b e r 5

SYMS et al. 545

Fig.3. Histopathologic slides al low (A) a n d high (B) power magnification of the spinal cord reveal a multifocal to diffuse, mild to moderate to severe meningoencephalomyelitis. The slides stained with hematoxylin a n d eosin demonstrate an intense neutrophilic meningitis (arrows) a n d marked perivasculitis (arrowhead). Slides are from d o g 7.

No neurologic sequelae were observed in dogs 8 through 16. The histopathologic condition of dogs 8 through 16 was normal. DISCUSSION

This study demonstrated that the use of a modified light source may make localization of a CSF fistula less arduous. Fluorescein-stained CSF was identified in all 16 dogs. The fistula in dog 16 was observed with a fluorescein concentration of 3.3 x 10 -s mg/ml, the lowest concentration used in the study. This demonstrates that the current usage of 0.25 ml of 5% fluorescein is well above the dose of intrathecal fluorescein needed to detect a CSF fistula. Humans have an approximate CSF volume of 140 ml. 11 A dosage of 0.25 ml of 5% fluorescein would result in a CSF concentration of fluorescein of 9.0 x 10 -5 mg/ml. This concentration could be obtained in a human using 4.6 x 10 -6 mg o~f fluorescein, equivalent to 4.6 x 10 -5 ml of 10% fluorescein. One drop of fluorescein placed in withdrawn CSF should be adequate to localize a CSF fistula using the modified light source.

Neurologic sequelae in humans have occurred at a concentration of fluorescein of 9.0 x 10-5 mg/ml. 6 In our experiment, the lowest concentration of fluorescein at which neurologic sequelae occurred was 8.3 x 10 -3 mg/ml. No neurologic sequelae were detected in dogs at concentrations of 5.0 x 10 -3 mg/ml and below. Most neurologic sequelae were immediate upon injection, and no animal had delayed onset of neurologic sequelae. The concentration of fluorescein that produced the neurologic sequelae was greater in this model than in humans. If the fluorescein is injected quickly, a region of higher concentration of fluorescein could be produced, causing neurologic sequelae. Alternatively, there may be a difference in sensitivities to fluorescein between humans and dogs. The histopathologic findings from necropsy specimens in dogs 1 through 7 revealed meningoencephalomyelitis, meningitis, and perivasculitis. Because of the rapidity of the development of the neurologic responses, we conclude that the sequelae were caused by the chemical irritation of the fluorescein. The lack of histopathologic changes in dogs 8 through 16

OtolaryngologyHead and Neck Surgery November 1997

546 SYMS et al.

supports this interpretation. Exposure to fluorescein concentrations of 8.3 x 10 -3 mg/ml and greater produced chemical irritation. This finding explains the histopathologic results. The histopathologic findings of this canine model could explain the neurologic sequelae that have been observed in humans. 4-6 The transient neurologic sequelae, observed in humans, are consistent with the acute meningoencephalomyelitis, meningitis, and perivasculitis produced by higher concentrations of fluorescein. Intrathecal fluorescein has been used clinically to detect C S F fistulae since its introduction by Kirchner in 1960. 3 Detection progressed from h e a d l i g h t visualization, p l e d g e t p l a c e m e n t in the nose, 12 and finally to direct endoscopic visualization.13,14 Although the use of filters may be viewed by some as unnecessary, we believe it adds a level of safety by enabling lower concentrations of fluorescein to be used. The technique of injection should include the introduction of the dye using the barbotage technique. 8 If these methods are incorporated into the endoscopic diagnosis of C S F fistulae, fluorescein can be used safely and accurately to identify the site o f the leak. We thank Donna Gabriel of the Department of Surgery at Tripler Army Medical Center for her assistance in the preparation of the manuscript and Adrian F. Van Dellen, DVM, LtCol, USAF, BSC, research pathologist, for his work on the necropsies.

REFERENCES 1. Spetzler RF, Zabramski JM. Cerebrospinal fluid fistulae: their management and repair. In: Youmans JR, editor. Neurological surgery. 3rd ed. Philadelphia: W.B. Saunders, 1990:2269. 2. Calcaterra TC. Diagnosis and management of ethmoid cerebrospinal rhinorrhea. Otolaryngol Ctin North Am 1985;18: 99-105. 3. Kirchner FR, Proud GO. Method for the identification and localization of cerebrospinal fluid, rhinorrhea and otorrhea. Laryngoscope 1960;70:921. 4. Mahaley MS, Odom GL. Complication following intrathecal injection of fluorescein. Case report. J Neurosurg 1966;25: 298-9. 5. Wallace JD, Weintraub MI, Mattson RH, Rosnagle R. Status epilepticus as a complication of intrathecal fluorescein. Case Report. J Neurosurg 1972;36:659-60. 6. Moseley JI, Carton CA, Stern WE. Spectrum of complicationsin the use of intrathecal fluorescein. J Neurosurg 1978;48:765-7. 7. Kirchner FR. Use of fluorescein for the diagnosis and localization of cerebrospinal fluid fistulas. Surgical Forum 1961;12:4068. 8. Hoerlein BE Canine neurology: diagnosis and treatment. 2nd ed. Philadelphia: W.B. Saunders, 1971:56l. 9. Nightingale PJ. Barbotage and spinal anaesthesia. Anaesthesia 1983;38:7-9. 10. Syms CA. Cerebrospinal fluid rhinorrhea. Ear Nose Throat J 1992;61:575-7. 11. Tourtellotte WW, Short RJ. Cerebrospinal fluid. In: Youmans JR, editor. Neurological surgery. 3rd ed. Philadelphia: W.B. Saunders, 1990:335. 12. Barrs DM, Kern EB. Use of intranasal pledgets for localization of cerebrospinal fluid rhinorrhea. Rhinology 1979;7:227-30. 13. Messerklinger W. Nasenendoskopie: Nachweis, Lokalisation und Differenfialdiagnose der nasalen Liquorrhoe. HNO 1972; 20:268-72. 14. Mattox DE, Kennedy DW. Endoscopic management of cerebrospinal fluid leaks and cephaloceles. Laryngoscope 1990;100: 857-62.

Craniofacial and Skull Base Surgery

The 2nd Congress on Craniofacial and Skull Base Surgery will be held Jan. 20-25, 1998, in Cancun, Mexico. CME credit is available. For further information contact Peggy M. Podboy, CMR FOCUS Meetings and Events, 625 S. George St.,, Mount Prospect, IL 60056-3915; phone (847)398-5821; fax (847)398-5822