- Email: [email protected]
Transorbital Sonographic Monitoring of Optic Nerve Diameter in Patients With Severe Brain Injury
Transorbital Sonographic Monitoring of Optic Nerve Diameter in Patients With Severe Brain Injury D. Karakitsos, T. Soldatos, A. Gouliamos, A. Armaganidis, J. Poularas, A. Kalogeromitros, J. Boletis, A. Kostakis, and A. Karabinis ABSTRACT Objective. We investigated whether alterations in the optic nerve diameter (OND) correlated with brain computed tomography (CT) imaging results among patients with brain injury and whether monitoring of OND could predict brain death. Patients and Methods. We enrolled 54 patients with brain injury (Glasgow Coma Scale ⬍ 8) and 53 controls. OND measurements were performed 3 mm posterior to the papillae by means of transorbital sonography. The severity of the injury was classified according to a semiquantitative CT neuroimaging scale (1 to 4). All patients underwent 3 repeated evaluations of OND combined with synchronous CT scans. Results. Twenty-two patients progressed to brain death, while 32 patients showed gradual clinical improvement. Upon admission, the patients showed significantly increased OND (4.84 ⫾ 1.2 mm) compared with the controls (3.49 ⫾ 1.1 mm; P ⬍ .001). The median intraobserver variation of OND was 0.2 mm (95% confidence intervals [CI]: 0.1– 0.7). The median interobserver variation of OND was 0.3 mm (95% CI: 0.1– 0.9). Alterations in the OND were significantly correlated with the neuroimaging scale on 3 repeated evaluations: r ⫽ .65, r ⫽ .70, and r ⫽ .73 (all P ⬍ .001). An OND greater than 5.9 mm (specificity ⫽ 65% and sensitivity ⫽ 74%; P ⬍ .01) and a 2.5 mm increased OND between repeated measurements (specificity ⫽ 70% and sensitivity ⫽ 81%; P ⬍ .01) were associated with a poor prognosis. Conclusions. Alterations in OND strongly correlated with neuroimaging results among patients with brain injury. However, monitoring of OND exhibited a low predictive value for brain death.
B
RAIN INJURY has been generally classified into mass lesion and diffuse injury, however, a more systematic understanding of its heterogeneous nature has evolved over the last decade.1 Elevated intracranial pressure (ICP) and consequent cerebral edema are both frequent manifestations of severe brain injury.2 The use of an intraventricular or subdural catheter remains the gold standard method to establish intracranial hypertension and evaluate patients with cerebral edema. However, this technique is highly invasive, exposing patients to the risks of central nervous system (CNS) infection and hemorrhage.3,4 Computed tomography (CT) of the CNS has been widely used to evaluate cerebral edema.2 Past studies have demonstrated a direct correlation between the severity of brain injury and CT imaging results.5 A semiquantitative classification scale of cerebral injury based on CT findings has
been suggested by Marshall et al.1 This scale describes the severity of diffuse injury and cerebral edema according to the status of the mesencephalic cisterns, the presence of a mass effect, and the degree of midline shift. It has been suggested to predict the probability of a fatal versus a nonfatal outcome.1 From the Departments of Intensive Care and Radiology-Imaging, General State Hospital of Athens, Athens, Greece (D.K., T.S., J.P., A.Kal., A.Kar.); Departments of Intensive Care and RadiologyImaging, Attikon University Hospital, Athens, Greece (A.G., A.A.); and Transplantation Center, Laiko University Hospital, Athens, Greece (J.B., A.K.). Address reprint requests to Andreas Karabinis, MD, PhD, 34 Mavromihaleon Street, Chalandri, 15233, Athens, Greece, Email: [email protected] and [email protected]
0041-1345/06/$–see front matter doi:10.1016/j.transproceed.2006.10.185
© 2006 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
3700
Transplantation Proceedings, 38, 3700 –3706 (2006)
TRANSORBITAL SONOGRAPHIC MONITORING OF OND
Other noninvasive methods to evaluate increased ICP are fontanometry and evoked potential studies; however, these methods require specialized equipment and experienced examiners.4 Interestingly, it has been documented that a direct communication exists between the perineural cerebrospinal fluid compartment and the craniospinal subarachnoid space, thus any change in ICP alters the optic nerve diameter (OND), particularly its anterior segment.6 This finding has been confirmed by postmortem studies which have revealed expanded OND in cases of fatal brain injuries.7 Direct visualization of the optic nerve and measurement of OND is possible via transorbital sonography, which has been employed previously to study alterations of OND in adult patients and in children with an elevated ICP.4,8 –13 Compared to ophthalmoscopy, which provides only a limited examination of the optic nerve at the level of the papilla, transorbital sonography allows the depiction of the optic nerve to a significant length. Also, the latter is not restricted by the presence of cataract and does not require administration of atropine. It is of note that in cases of raised ICP, the OND increases initially leading to papilledema; thus, transorbital sonography may depict cases of increased ICP earlier than ophthalmoscopy.4 We investigated whether alterations in the OND as assessed by means of transorbital sonography correlated with alterations in the anatomical state of the brain as depicted by synchronous brain CT examinations in a cohort of patients with severe brain injury. Furthermore, we analyzed whether sonographic monitoring of OND predicted progression of injury toward tamponade and therefore brain death (BD).
PATIENTS AND METHODS Patients This study was performed from January 2003 to March 2006 in the intensive care unit (ICU). Fifty-four patients with severe brain injury (Glasgow Coma Scale [GCS] ⬍ 8) and 53 patients without brain injury (controls) participated in this study. We excluded patients with orbitofacial trauma, history of glaucoma, or known disease of the optic nerve (inflammation, tumor). All patients were continuously monitored for systemic blood pressure, heart rate, PaO2, and PaCO2, to maintain steady state conditions and prevent hypotension (systolic blood pressure ⬎ 110 mm Hg), bradycardia (heart rate ⬎ 60 beats/min), or hypoxia (SpaO2 ⬍ 95%). For all patients, PaCO2 was maintained at 33 to 35 mm Hg throughout the study. Clinical BD was diagnosed according to the following criteria: deep irreversible coma, absence of brain stem reflexes, flat electroencephalogram, and a positive apnea test in a normothermic nondrugged patient.14,15 According to Hellenic state law, a confirmatory test (standard angiography) was required to document cessation of cerebral blood flow and to declare a patient brain dead. In our department, following a clinical diagnosis of BD, contrast angiography as well as transcranial Doppler sonography were performed to establish the diagnosis of BD as described in detail elsewhere.16 Thereafter, the family’s consent was obtained for organ donation by a committee consisting of an intensive care consultant, a neurologist, a cardiologist, a neurosurgeon, and an anesthesiologist. In this series, 12 patients became organ and tissue donors. The study conformed with the principles
3701 outlined in the Declaration of Helsinki and was approved by the Institutional Ethics Committee.
Methods All patients underwent CT imaging of the CNS for the evaluation of brain injury (Fig 1A). Based on the neuroimaging results, each patient was categorized with respective to a semiquantitative scale (1 to 4), as previously described1 (Table 1). Transorbital sonography of the optic nerve was performed, using an HDI 3500 (ATL, Philips, Bothell, USA) equipped with a 7.5 MHz linear transducer. Soft placement of the probe on the upper temporal eyelid and proper adjustment of the insonation angle provided an axial view of the orbit and the structures of the retrobulbar area. An insonation depth of 6.2 mm was selected and individual adjustments of the ultrasound gain and the output sound intensity were made for each patient to provide the highest level of contrast. After the depiction of the optic nerve in the axial plane, cursors were placed on the outer contours of the dural sheath, 3 mm posterior to the papillae (Fig 1B). The OND was calculated as the horizontal distance between the 2 cursors. Two independent observers blind to the subjects’ identities performed all sonographic examinations. Three measurements were taken for each optic nerve during each measurement session and average values were used in the statistical analysis. The OND registered was the average value obtained from measurements of both eyes, despite the fact that previous studies have shown the presence of intraocular symmetry between the ONDs of fellow eyes.17,18 We calculated the difference from the mean value for each eye of each single measurement for each observer. Median values of this intraobserver variation were then calculated for the total study population. Accordingly, median values for the differences between the mean values from each observer (interobserver variation) were also determined.19,20 Upon admission each patient was clinically assessed by means of GCS. The acute physiology and chronic health evaluation (APACHE). 11 score was calculated with the values obtained within the first 24 hours of ICU admission. Thereafter, sonographic evaluation of the OND and CT evaluation of the brain injury were performed. During the study period, 3 sonographic measurements of the OND were combined with synchronous brain CT examinations for each patient.
Statistical Analysis Summary data are expressed as mean values ⫾ standard deviations. One factor repeated measures ANOVA model was used to compare each variable during the observation period. Pair-wise multiple comparisons were performed using the Tukey critical difference method. Spearman’s correlation test was used to evaluate correlations between measurements of the OND and the semiquantitative scale of brain injury (1 to 4). A receiver operating characteristic (ROC) curve analysis was performed to obtain cutoff levels of the OND and its alterations to classify patients as high versus low risk for BD by calculating the respective areas under the curve.21 Values of P ⬍ .05 were considered significant. All tests were 2-sided and analysis was performed using the SPSS software, version 11 (SPSS Inc).
RESULTS
The baseline characteristics of the study population are presented in Table 2. The mean period of ICU hospitaliza-
3702
KARAKITSOS, SOLDATOS, GOULIAMOS ET AL
Fig 1. (A) Admission brain CT scan of a patient with intracerebral and subarachnoid hemorrhage showing a midline shift ⬎5 mm (arrow) and a nonevacuated lesion ⬍ 25 mL, findings that classify this case as level 4 (severe brain edema) on the neuroimaging scale. (B) Transorbital sonography of the same patient. Axial image of the postocular area demonstrates the optic nerve sheath complex as a hypoechoic stripe in between echogenic fat. Cursors were placed 3 mm behind the globe on the outer contours of the nerve sheath (arrows).
tion for patients with severe brain injury was 38 ⫾ 47 days. Among the 54 patients with brain injury, 22 progressed to BD, while 32 patients showed gradual clinical improvement and were finally discharged. There were no significant differences in sex, body surface area, and APACHE II score between patients and controls; however, patients were significantly younger compared with controls (Table 2). The OND was significantly increased
upon admission among patients with brain injury (4.84 ⫾ 1.2 mm) compared with controls (3.49 ⫾ 1.1 mm P ⬍ .001; Table 2). Moreover, at baseline the OND inversely correlated with age (r ⫽ ⫺.34, P ⬍ .01) among all subjects. It is of note that control subjects with adult respiratory distress syndrome (ARDS) showed significantly increased (P ⬍ .05) OND (3.6 ⫾ 0.4 mm) compared with other control subjects (3.38 ⫾ 1.3 mm).
TRANSORBITAL SONOGRAPHIC MONITORING OF OND Table 1. Classification of Brain Injury Based on CT Scan Findings Brain Injury Scale
CT Scan Findings
1 2 3 4 4a 4b
Normal CT scan (no visible pathology) Cisterns present; midline shift 0–5 mm Cisterns compressed or absent; midline shift 0–5 mm Midline shift ⬎5 mm Any surgically evacuated mass lesion Lesion ⬎25 mL not surgically evacuated
The median intraobserver variation of OND was 0.2 mm (95% confidence intervals [CI]: 0.1– 0.7). The median interobserver variation of OND was 0.3 mm (95% CI: 0.1– 0.9). Furthermore, all patients with brain injury underwent 3 repeated sonographic OND measurements along with synchronous clinical and brain CT imaging evaluations. At the second evaluation (12 ⫾ 3.5 days from baseline), the OND was 5.1 ⫾ 1.2 mm, the GCS was 7.3 ⫾ 3, and the diffuse injury scale was 2 ⫾ 1.9, all significantly altered compared with the baseline values (P ⬍ .005). At the third evaluation (35 ⫾ 6.2 days from baseline), the OND was 5.3 ⫾ 1.3 mm, the GCS was 9.3 ⫾ 4.5, and the neuroimaging scale was 1.8 ⫾ 2.6, all significantly altered compared with the baseline values (P ⬍ .005). The sonographic OND measurements significantly correlated (P ⬍ .001) with the neuroimaging scale in all repeated evaluations (Fig 2). The ROC curves illustrate the relationship between sensitivity and specificity to determine the predictive value for a poor prognosis of OND measurements. Using the ROC curves, we determined that the threshold values associated with a poor prognosis were an absolute value of OND greater than 5.9 mm (area under the curve ⫽ 0.805 [95% CI ⫽ 0.768 – 0.911]; specificity ⫽ 65% and sensitivity ⫽ 74%; P ⬍ .01) and an increase of more than 2.5 mm in the OND among repeated measurements (area under the curve ⫽ 0.832 [95% CI ⫽ 0.794 – 0.926]; specificity ⫽ 70% and sensitivity ⫽ 81%; P ⬍ .01) (Fig 3). DISCUSSION
The hypothesis of this study was that the severity of cerebral edema in patients with brain injury could be estimated by sonographic evaluation of OND. In most cases, the clinical examination was not sufficient; thus there was a necessity for other diagnostic methods. Until recently, reliable assessment of cerebral edema has been possible by means of invasive methods and by CT imaging. In this study, patients with brain injury and consequent cerebral edema exhibited an increased OND compared with subjects without any brain injury. Interestingly, control subjects with ARDS showed increased OND compared with other control subjects. The enlargement of the optic nerve is believed to indicate a connection between the subarachnoid spaces of the brain and the perioptic nerve sheath. Furthermore, this possibly explains the intrathecal enhancement of the perioptic nerve
3703
space during myelography or cisternography.22 In a study of fresh cadavers, Liu and Kahn23 performed saline infusions through a ventriculostomy to achieve various levels of ICP. They observed a linear relationship between the latter and the subarachnoid pressure of the optic nerve as recorded through an orbitotomy. They suggested that the ICP induces an anterior flow of cerebrospinal fluid which fills the bulbous portion of the nerve. The latter is then squeezed by the globe’s movement so that the direction of the flow is reversed, and the circulation of the cerebrospinal fluid is completed along the nerve. However, the above study lacked many factors which possibly contribute to the phenomenon of optic nerve enlargement, such as tissue elasticity, blood-brain barrier, respiration, and continuous production and resorption of cerebrospinal fluid. Furthermore, Hansen and Helmke24 used transorbital sonography to investigate the OND response to pressure during cerebrospinal fluid absorption studies in patients undergoing neurological testing. They observed that in all cases the OND changes demonstrated covariance with the alteration of lumbar cerebrospinal fluid pressure. Additionally, the OND changes were completely reversible during the infusion tests. This observation confirmed that the optic nerve sheath has adequate elasticity to allow detectable expansion in cases of raised ICP and underlies the equilibration of the cerebrospinal fluid pressure between the orbital and cranial cavities.24 Other studies have demonstrated the presence of lymphatic capillaries in the dura of the optic nerve which may serve as an alternative pathway for the drainage of cerebrospinal fluid.25–28 However, this outflow pathway may not be adequate for the absorption of enough cerebrospinal fluid in cases of intracranial hypertension.29 The present results showed that OND measurements Table 2. Baseline Characteristics of the Study Population
Parameters
Age (y) Sex (M/F) BSA (m2) GCS (1–14) APACHE II score Diagnosis upon admission
Neuroimaging brain injury scale upon admission (1–4) Brain death OND upon admission (mm)
Subjects With Brain Injury (n ⫽ 54)
Subjects Without Brain Injury (Controls; n ⫽ 53)
40 ⫾ 18.6* 28/26 20 ⫾ 4.3 5.7 ⫾ 2.1 20 ⫾ 3.5 Head injury: 25 (46.2%)
49 ⫾ 16.8 26/27 20 ⫾ 5.1 — 19 ⫾ 5.1 Sepsis: 18 (33.9%)
CEH: 15 (27.7%) SUH: 14 (25.9%) 2.1 ⫾ 1.1
ARDS: 14 (26.4%) Trauma: 14 (26.4%) Burn: 7 (13.2%) —
22 (40.7%) 4.84 ⫾ 1.2*
— 3.49 ⫾ 1.1
BSA ⫽ body surface area; GCS ⫽ Glasgow Coma Scale; APACHE II score-acute physiology and chronic health evaluation score; CEH ⫽ cerebral hemorrhage; SUH ⫽ subarachnoid hemorrhage; ARDS ⫽ adult respiratory distress syndrome. *P ⬍ .001 by ANOVA.
3704
KARAKITSOS, SOLDATOS, GOULIAMOS ET AL
Fig 2. OND measurements plotted versus the semiquantitative CT (neuroimaging) scale (1 to 4) in patients with severe brain injury (n ⫽ 54) at baseline (A), at second evaluation (B), and at third evaluation (C).
were significantly correlated with the CT imaging results in patients with brain injury and edema. Although occasional cases of patients with intracranial hypertension and enlarged OND have been described,9,10 few studies exist that have assessed this correlation. Blaivas et al8 examined adult patients with elevated ICP values from a variety of causes, suggesting that an OND above 5 mm is indicative of
intracranial hypertension. In a similar study, Brzezinska and Schumacher11 demonstrated significantly wide ONDs in patients with raised ICP and suggested that the upper normal limit of the optic nerve was 4.1 mm. Helmke and Hansen12 examined children with elevated ICP due to head trauma or a metabolic disorder, concluding that the OND is definitely enlarged when exceeding 5 mm in children older
TRANSORBITAL SONOGRAPHIC MONITORING OF OND
3705
Fig 3. (A) Receiver operating characteristic curve showing the relationship between sensitivity and specificity to determine the predictive value of an increase in the OND (the cutoff value is 2.5 mm) for brain death. The diagonal line is the nondiscriminant curve (B) Receiver operating characteristic curve showing the relationship between sensitivity and specificity to determine the predictive value of the OND (the cutoff value is 5.9 mm) for brain death. The diagonal line is the nondiscriminant line.
than age 4. An analogous study in a pediatric population was conducted by Newman et al,13 who reported an upper normal limit of 4.0 mm for patients less than 1 year old and 4.5 mm for older children. They suggested that higher OND values were strongly indicative of raised ICP. Finally, Malayeri et al4 examined children with elevated ICP observing that all patients either younger or older than age 4 exhibited OND above 4.55 mm. This study suggested that patients with brain injury who exhibited OND ⱖ 5.9 mm as well as an increased OND during sonographic monitoring of more than 2.5 mm had a high probability to progress toward brain tamponade and BD. However, the specificity and sensitivity of the above findings were rather low. Therefore, one cannot suggest that sonographic monitoring of the OND is a good candidate to predict final outcome in patients with brain injury. It is of note that alterations in the OND had a close relationship with brain CT results during the study period which may be of clinical value for the diagnosis of cerebral edema. Compared with CT, which is a highly accurate method for the evaluation of brain injury, transorbital sonography offers the significant advantages of low cost, wide availability, and rapid bedside performance. Moreover, this method is noninvasive and does not expose the patient to radiation. In this study, we used a neuroimaging scale instead of an intraventricular or subdural catheter to evaluate cerebral edema. Therefore, small variations of the ICP cannot be detected unless they sum to levels which alter the anatomical state of the brain within limits depicted by means of CT. Also, the pathophysiology of brain injury is quite complex as brain edema progresses toward brain tamponade and
therefore continuous monitoring of a variety of metabolic and circulatory parameters may be necessary.16 Transorbital sonography of the OND offers the advantage of continuous monitoring and may assist in the evaluation of brain injury, provided that examinations are performed using an accurate technique. Age and glaucoma are factors that should be taken into account when measuring the OND.31 We found an inverse relationship between age and OND which is in accord with the findings of Beatty et al.31 Also, optic nerve enlargement can occur due to secondary involvement from a variety of orbital and systematic abnormalities, such as tumor, inflammation, Grave’s disease, sarcoidosis, pseudotumor, metastasis, and hemorrhage in and around the optic nerve complex as well as hydrops from an extrinsic tumor.22 The present data showed that transorbital sonography has high intra- and interobserver reproducibility as others have also reported.17,31 However, this method relies heavily on the skill and diligence of the observer. Common technical pitfalls include: 1) inadequate depiction of the optic nerve in the transverse plane, 2) inaccurate designation of the OND contours, and 3) erroneous placement of the cursors.17 Finally, OND measurements should be performed 3 mm posterior to the papillae since the respective segment of the optic nerve features the highest level of distension and provides optimum sonographic contrast between the echogenic fat and the hypoechoic optic nerve.6,13,30 In conclusion, transorbital sonography of the OND is a highly reproducible method and may be used to evaluate brain injury and consequently cerebral edema. Alterations
3706
in the OND showed strong correlations with synchronous CT imaging findings in patients with severe brain injury. However, the sonographic monitoring of OND in the above patients showed a low specificity and sensitivity to identify patients at high risk for brain tamponade and consequently brain death. Further studies with larger numbers of patients are clearly required to determine whether sonographic monitoring of OND should be adopted more widely as a diagnostic and prognostic tool in neurocritical care. REFERENCES 1. Marshall LF, Marshall SB, Klauber MR, et al: The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 9:287, 1992 2. Barr R, Gean A: Craniofacial trauma. In Brant W, Helms C (eds): Fundamentals of Diagnostic Radiology, 2nd Ed. Philadelphia: Lippincott Williams & Wilkins; 1999, p 49 3. Czosnyka M, Pickard JD: Monitoring and interpretation of intracranial pressure. J Neurol Neurosurg Psychiatry 75:813, 2004 4. Malayeri AA, Bavarian S, Mehdizadeh M: Sonographic evaluation of optic nerve diameter in children with raised intracranial pressure. J Ultrasound Med 24:143, 2005 5. Gennarelli TA, Spielman GM, Langfitt TW, et al: Influence of the type of intracranial lesion on outcome from severe head injury. J Neurosurg 56:26, 1992 6. Hansen HC, Helmke K: The subarachnoid space surrounding the optic nerves. An ultrasound study of the optic nerve sheath. Surg Radiol Anat 18:323, 1996 7. Muller PJ, Deck JHN: Intraocular and optic nerve sheath hemorrhage in cases of sudden intracranial hypertension. J Neurosurg 41:160, 1974 8. Blaivas M, Theodoro D, Sierzenski PR: Elevated intracranial pressure detected by bedside emergency ultrasonography of optic nerve sheath. Acad Emerg Med 45:336, 2005 9. Tsung JW, Blaivas M, Cooper A, et al: A rapid noninvasive method of detecting elevated intracranial pressure using bedside ocular ultrasound: application to 3 cases of head trauma in the pediatric emergency department. Pediatr Emerg Care 21:94, 2005 10. Galetta S, Byrne SF, Smith JL: Echographic correlation of optic nerve sheath size and cerebrospinal fluid pressure. J Clin Neuroophthalmol 9:79, 1989 11. Brzezinska R, Schumacher R: Diagnosis of elevated intracranial pressure in children with shunt under special consideration of transglobe sonography of the optic nerve. Ultraschall Med 23:325, 2002 12. Helmke H, Hansen HC: Fundamentals of transorbital sonographic evaluation of optic nerve sheath expansion under intracranial hypertension. II. Patient study. Pediatr Radiol 26:706, 1996 13. Newman WD, Hollman AS, Dutton GN, et al: Measurement of optic nerve sheath diameter by ultrasound: a means of detecting acute raised intracranial pressure in hydrocephalus. Br J Ophthalmol 86:1109, 2002
KARAKITSOS, SOLDATOS, GOULIAMOS ET AL 14. Report of the medical consultants on the diagnosis of death to the President’s commission for the study of ethical problems in medicine and biomedical and behavioural research. Guidelines for the determination of death. JAMA 246:2184, 1981 15. Wijdicks EFM: The diagnosis of brain death. N Engl J Med 344:1215, 2001 16. Poularas J, Karakitsos D, Kouraklis G, et al: Comparison between transcranial colour Doppler ultrasonography and angiography in the confirmation of brain death. Transplant Proc (in press) 17. Ballantyne SA, O’Neill G, Hamilton R, et al: Observer variation in the sonographic measurement of optic nerve sheath diameter in normal adults. Eur J Ultrasound 15:145, 2002 18. Karim S, Clark RA, Poukens V, et al: Demonstration of systematic variation in human intraorbital optic nerve size by quantitative magnetic resonance imaging and histology. Invest Ophthalmol Vis Sci 45:1047, 2004 19. Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307, 1986 20. Bland JM, Altman DG: Correlation, regression, and repeated data. Br Med J 308:896, 1994 21. Sweets J: Measuring the accuracy of diagnosis system. Science 240:1285, 1988 22. Rothman MI, Zoarski GH: The Orbit. In Sutton D (ed): Textbook of Radiology and Imaging, 7th Ed. London: Churchill Livingstone; 2003, p 1573 23. Liu D, Kahn M: Measurement and relationship of subarachnoid pressure of the optic nerve to intracranial pressures in fresh cadavers. Am J Ophthalmol 116:548, 1993 24. Hansen HC, Helmke K: Validation of the optic nerve sheath response to changing cerebrospinal fluid pressure: ultrasound findings during intrathecal infusion tests. J Neurosurg 87:34, 1997 25. Killer HE, Laeng HR, Groscurth P: Lymphatic capillaries in the meninges of the human optic nerve. J Neuroophthalmol 19:222, 1999 26. Gausas RE, Gonnering RS, Lemke BN, et al: Identification of human orbital lymphatics. Ophthal Plast Reconstr Surg 15:252, 1999 27. Oehmichen M, Gruninger H, Wietholter H, et al: Lymphatic efflux of intracerebrally injected cells. Acta Neuropathol (Berl) 45:61, 1979 28. Brinker T, Ludemann W, Berens von Rautenfeld D, et al: Dynamic properties of lymphatic pathways for the absorption of cerebrospinal fluid. Acta Neuropathol (Berl) 94:493, 1997 29. Killer HE, Laeng HR, Flammer J, et al: Architecure of arachnoid trabeculae, pillars and septa in the subarachnoid space of the human optic nerve: anatomy and clinical considerations. Br J Ophthalmol 87:777, 2003 30. Helmke H, Hansen HC: Fundamentals of transorbital sonographic evaluation of optic nerve sheath expansion under intracranial hypertension. I. Experimental study. Pediatr Radiol 26:701, 1996 31. Beatty S, Good PA, McLaughlin J, et al: Echographic measurements of the retrobulbar optic nerve in normal and glaucomatous eyes. Br J Ophthalmol 82:43, 1998
B
RAIN INJURY has been generally classified into mass lesion and diffuse injury, however, a more systematic understanding of its heterogeneous nature has evolved over the last decade.1 Elevated intracranial pressure (ICP) and consequent cerebral edema are both frequent manifestations of severe brain injury.2 The use of an intraventricular or subdural catheter remains the gold standard method to establish intracranial hypertension and evaluate patients with cerebral edema. However, this technique is highly invasive, exposing patients to the risks of central nervous system (CNS) infection and hemorrhage.3,4 Computed tomography (CT) of the CNS has been widely used to evaluate cerebral edema.2 Past studies have demonstrated a direct correlation between the severity of brain injury and CT imaging results.5 A semiquantitative classification scale of cerebral injury based on CT findings has
been suggested by Marshall et al.1 This scale describes the severity of diffuse injury and cerebral edema according to the status of the mesencephalic cisterns, the presence of a mass effect, and the degree of midline shift. It has been suggested to predict the probability of a fatal versus a nonfatal outcome.1 From the Departments of Intensive Care and Radiology-Imaging, General State Hospital of Athens, Athens, Greece (D.K., T.S., J.P., A.Kal., A.Kar.); Departments of Intensive Care and RadiologyImaging, Attikon University Hospital, Athens, Greece (A.G., A.A.); and Transplantation Center, Laiko University Hospital, Athens, Greece (J.B., A.K.). Address reprint requests to Andreas Karabinis, MD, PhD, 34 Mavromihaleon Street, Chalandri, 15233, Athens, Greece, Email: [email protected] and [email protected]
0041-1345/06/$–see front matter doi:10.1016/j.transproceed.2006.10.185
© 2006 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
3700
Transplantation Proceedings, 38, 3700 –3706 (2006)
TRANSORBITAL SONOGRAPHIC MONITORING OF OND
Other noninvasive methods to evaluate increased ICP are fontanometry and evoked potential studies; however, these methods require specialized equipment and experienced examiners.4 Interestingly, it has been documented that a direct communication exists between the perineural cerebrospinal fluid compartment and the craniospinal subarachnoid space, thus any change in ICP alters the optic nerve diameter (OND), particularly its anterior segment.6 This finding has been confirmed by postmortem studies which have revealed expanded OND in cases of fatal brain injuries.7 Direct visualization of the optic nerve and measurement of OND is possible via transorbital sonography, which has been employed previously to study alterations of OND in adult patients and in children with an elevated ICP.4,8 –13 Compared to ophthalmoscopy, which provides only a limited examination of the optic nerve at the level of the papilla, transorbital sonography allows the depiction of the optic nerve to a significant length. Also, the latter is not restricted by the presence of cataract and does not require administration of atropine. It is of note that in cases of raised ICP, the OND increases initially leading to papilledema; thus, transorbital sonography may depict cases of increased ICP earlier than ophthalmoscopy.4 We investigated whether alterations in the OND as assessed by means of transorbital sonography correlated with alterations in the anatomical state of the brain as depicted by synchronous brain CT examinations in a cohort of patients with severe brain injury. Furthermore, we analyzed whether sonographic monitoring of OND predicted progression of injury toward tamponade and therefore brain death (BD).
PATIENTS AND METHODS Patients This study was performed from January 2003 to March 2006 in the intensive care unit (ICU). Fifty-four patients with severe brain injury (Glasgow Coma Scale [GCS] ⬍ 8) and 53 patients without brain injury (controls) participated in this study. We excluded patients with orbitofacial trauma, history of glaucoma, or known disease of the optic nerve (inflammation, tumor). All patients were continuously monitored for systemic blood pressure, heart rate, PaO2, and PaCO2, to maintain steady state conditions and prevent hypotension (systolic blood pressure ⬎ 110 mm Hg), bradycardia (heart rate ⬎ 60 beats/min), or hypoxia (SpaO2 ⬍ 95%). For all patients, PaCO2 was maintained at 33 to 35 mm Hg throughout the study. Clinical BD was diagnosed according to the following criteria: deep irreversible coma, absence of brain stem reflexes, flat electroencephalogram, and a positive apnea test in a normothermic nondrugged patient.14,15 According to Hellenic state law, a confirmatory test (standard angiography) was required to document cessation of cerebral blood flow and to declare a patient brain dead. In our department, following a clinical diagnosis of BD, contrast angiography as well as transcranial Doppler sonography were performed to establish the diagnosis of BD as described in detail elsewhere.16 Thereafter, the family’s consent was obtained for organ donation by a committee consisting of an intensive care consultant, a neurologist, a cardiologist, a neurosurgeon, and an anesthesiologist. In this series, 12 patients became organ and tissue donors. The study conformed with the principles
3701 outlined in the Declaration of Helsinki and was approved by the Institutional Ethics Committee.
Methods All patients underwent CT imaging of the CNS for the evaluation of brain injury (Fig 1A). Based on the neuroimaging results, each patient was categorized with respective to a semiquantitative scale (1 to 4), as previously described1 (Table 1). Transorbital sonography of the optic nerve was performed, using an HDI 3500 (ATL, Philips, Bothell, USA) equipped with a 7.5 MHz linear transducer. Soft placement of the probe on the upper temporal eyelid and proper adjustment of the insonation angle provided an axial view of the orbit and the structures of the retrobulbar area. An insonation depth of 6.2 mm was selected and individual adjustments of the ultrasound gain and the output sound intensity were made for each patient to provide the highest level of contrast. After the depiction of the optic nerve in the axial plane, cursors were placed on the outer contours of the dural sheath, 3 mm posterior to the papillae (Fig 1B). The OND was calculated as the horizontal distance between the 2 cursors. Two independent observers blind to the subjects’ identities performed all sonographic examinations. Three measurements were taken for each optic nerve during each measurement session and average values were used in the statistical analysis. The OND registered was the average value obtained from measurements of both eyes, despite the fact that previous studies have shown the presence of intraocular symmetry between the ONDs of fellow eyes.17,18 We calculated the difference from the mean value for each eye of each single measurement for each observer. Median values of this intraobserver variation were then calculated for the total study population. Accordingly, median values for the differences between the mean values from each observer (interobserver variation) were also determined.19,20 Upon admission each patient was clinically assessed by means of GCS. The acute physiology and chronic health evaluation (APACHE). 11 score was calculated with the values obtained within the first 24 hours of ICU admission. Thereafter, sonographic evaluation of the OND and CT evaluation of the brain injury were performed. During the study period, 3 sonographic measurements of the OND were combined with synchronous brain CT examinations for each patient.
Statistical Analysis Summary data are expressed as mean values ⫾ standard deviations. One factor repeated measures ANOVA model was used to compare each variable during the observation period. Pair-wise multiple comparisons were performed using the Tukey critical difference method. Spearman’s correlation test was used to evaluate correlations between measurements of the OND and the semiquantitative scale of brain injury (1 to 4). A receiver operating characteristic (ROC) curve analysis was performed to obtain cutoff levels of the OND and its alterations to classify patients as high versus low risk for BD by calculating the respective areas under the curve.21 Values of P ⬍ .05 were considered significant. All tests were 2-sided and analysis was performed using the SPSS software, version 11 (SPSS Inc).
RESULTS
The baseline characteristics of the study population are presented in Table 2. The mean period of ICU hospitaliza-
3702
KARAKITSOS, SOLDATOS, GOULIAMOS ET AL
Fig 1. (A) Admission brain CT scan of a patient with intracerebral and subarachnoid hemorrhage showing a midline shift ⬎5 mm (arrow) and a nonevacuated lesion ⬍ 25 mL, findings that classify this case as level 4 (severe brain edema) on the neuroimaging scale. (B) Transorbital sonography of the same patient. Axial image of the postocular area demonstrates the optic nerve sheath complex as a hypoechoic stripe in between echogenic fat. Cursors were placed 3 mm behind the globe on the outer contours of the nerve sheath (arrows).
tion for patients with severe brain injury was 38 ⫾ 47 days. Among the 54 patients with brain injury, 22 progressed to BD, while 32 patients showed gradual clinical improvement and were finally discharged. There were no significant differences in sex, body surface area, and APACHE II score between patients and controls; however, patients were significantly younger compared with controls (Table 2). The OND was significantly increased
upon admission among patients with brain injury (4.84 ⫾ 1.2 mm) compared with controls (3.49 ⫾ 1.1 mm P ⬍ .001; Table 2). Moreover, at baseline the OND inversely correlated with age (r ⫽ ⫺.34, P ⬍ .01) among all subjects. It is of note that control subjects with adult respiratory distress syndrome (ARDS) showed significantly increased (P ⬍ .05) OND (3.6 ⫾ 0.4 mm) compared with other control subjects (3.38 ⫾ 1.3 mm).
TRANSORBITAL SONOGRAPHIC MONITORING OF OND Table 1. Classification of Brain Injury Based on CT Scan Findings Brain Injury Scale
CT Scan Findings
1 2 3 4 4a 4b
Normal CT scan (no visible pathology) Cisterns present; midline shift 0–5 mm Cisterns compressed or absent; midline shift 0–5 mm Midline shift ⬎5 mm Any surgically evacuated mass lesion Lesion ⬎25 mL not surgically evacuated
The median intraobserver variation of OND was 0.2 mm (95% confidence intervals [CI]: 0.1– 0.7). The median interobserver variation of OND was 0.3 mm (95% CI: 0.1– 0.9). Furthermore, all patients with brain injury underwent 3 repeated sonographic OND measurements along with synchronous clinical and brain CT imaging evaluations. At the second evaluation (12 ⫾ 3.5 days from baseline), the OND was 5.1 ⫾ 1.2 mm, the GCS was 7.3 ⫾ 3, and the diffuse injury scale was 2 ⫾ 1.9, all significantly altered compared with the baseline values (P ⬍ .005). At the third evaluation (35 ⫾ 6.2 days from baseline), the OND was 5.3 ⫾ 1.3 mm, the GCS was 9.3 ⫾ 4.5, and the neuroimaging scale was 1.8 ⫾ 2.6, all significantly altered compared with the baseline values (P ⬍ .005). The sonographic OND measurements significantly correlated (P ⬍ .001) with the neuroimaging scale in all repeated evaluations (Fig 2). The ROC curves illustrate the relationship between sensitivity and specificity to determine the predictive value for a poor prognosis of OND measurements. Using the ROC curves, we determined that the threshold values associated with a poor prognosis were an absolute value of OND greater than 5.9 mm (area under the curve ⫽ 0.805 [95% CI ⫽ 0.768 – 0.911]; specificity ⫽ 65% and sensitivity ⫽ 74%; P ⬍ .01) and an increase of more than 2.5 mm in the OND among repeated measurements (area under the curve ⫽ 0.832 [95% CI ⫽ 0.794 – 0.926]; specificity ⫽ 70% and sensitivity ⫽ 81%; P ⬍ .01) (Fig 3). DISCUSSION
The hypothesis of this study was that the severity of cerebral edema in patients with brain injury could be estimated by sonographic evaluation of OND. In most cases, the clinical examination was not sufficient; thus there was a necessity for other diagnostic methods. Until recently, reliable assessment of cerebral edema has been possible by means of invasive methods and by CT imaging. In this study, patients with brain injury and consequent cerebral edema exhibited an increased OND compared with subjects without any brain injury. Interestingly, control subjects with ARDS showed increased OND compared with other control subjects. The enlargement of the optic nerve is believed to indicate a connection between the subarachnoid spaces of the brain and the perioptic nerve sheath. Furthermore, this possibly explains the intrathecal enhancement of the perioptic nerve
3703
space during myelography or cisternography.22 In a study of fresh cadavers, Liu and Kahn23 performed saline infusions through a ventriculostomy to achieve various levels of ICP. They observed a linear relationship between the latter and the subarachnoid pressure of the optic nerve as recorded through an orbitotomy. They suggested that the ICP induces an anterior flow of cerebrospinal fluid which fills the bulbous portion of the nerve. The latter is then squeezed by the globe’s movement so that the direction of the flow is reversed, and the circulation of the cerebrospinal fluid is completed along the nerve. However, the above study lacked many factors which possibly contribute to the phenomenon of optic nerve enlargement, such as tissue elasticity, blood-brain barrier, respiration, and continuous production and resorption of cerebrospinal fluid. Furthermore, Hansen and Helmke24 used transorbital sonography to investigate the OND response to pressure during cerebrospinal fluid absorption studies in patients undergoing neurological testing. They observed that in all cases the OND changes demonstrated covariance with the alteration of lumbar cerebrospinal fluid pressure. Additionally, the OND changes were completely reversible during the infusion tests. This observation confirmed that the optic nerve sheath has adequate elasticity to allow detectable expansion in cases of raised ICP and underlies the equilibration of the cerebrospinal fluid pressure between the orbital and cranial cavities.24 Other studies have demonstrated the presence of lymphatic capillaries in the dura of the optic nerve which may serve as an alternative pathway for the drainage of cerebrospinal fluid.25–28 However, this outflow pathway may not be adequate for the absorption of enough cerebrospinal fluid in cases of intracranial hypertension.29 The present results showed that OND measurements Table 2. Baseline Characteristics of the Study Population
Parameters
Age (y) Sex (M/F) BSA (m2) GCS (1–14) APACHE II score Diagnosis upon admission
Neuroimaging brain injury scale upon admission (1–4) Brain death OND upon admission (mm)
Subjects With Brain Injury (n ⫽ 54)
Subjects Without Brain Injury (Controls; n ⫽ 53)
40 ⫾ 18.6* 28/26 20 ⫾ 4.3 5.7 ⫾ 2.1 20 ⫾ 3.5 Head injury: 25 (46.2%)
49 ⫾ 16.8 26/27 20 ⫾ 5.1 — 19 ⫾ 5.1 Sepsis: 18 (33.9%)
CEH: 15 (27.7%) SUH: 14 (25.9%) 2.1 ⫾ 1.1
ARDS: 14 (26.4%) Trauma: 14 (26.4%) Burn: 7 (13.2%) —
22 (40.7%) 4.84 ⫾ 1.2*
— 3.49 ⫾ 1.1
BSA ⫽ body surface area; GCS ⫽ Glasgow Coma Scale; APACHE II score-acute physiology and chronic health evaluation score; CEH ⫽ cerebral hemorrhage; SUH ⫽ subarachnoid hemorrhage; ARDS ⫽ adult respiratory distress syndrome. *P ⬍ .001 by ANOVA.
3704
KARAKITSOS, SOLDATOS, GOULIAMOS ET AL
Fig 2. OND measurements plotted versus the semiquantitative CT (neuroimaging) scale (1 to 4) in patients with severe brain injury (n ⫽ 54) at baseline (A), at second evaluation (B), and at third evaluation (C).
were significantly correlated with the CT imaging results in patients with brain injury and edema. Although occasional cases of patients with intracranial hypertension and enlarged OND have been described,9,10 few studies exist that have assessed this correlation. Blaivas et al8 examined adult patients with elevated ICP values from a variety of causes, suggesting that an OND above 5 mm is indicative of
intracranial hypertension. In a similar study, Brzezinska and Schumacher11 demonstrated significantly wide ONDs in patients with raised ICP and suggested that the upper normal limit of the optic nerve was 4.1 mm. Helmke and Hansen12 examined children with elevated ICP due to head trauma or a metabolic disorder, concluding that the OND is definitely enlarged when exceeding 5 mm in children older
TRANSORBITAL SONOGRAPHIC MONITORING OF OND
3705
Fig 3. (A) Receiver operating characteristic curve showing the relationship between sensitivity and specificity to determine the predictive value of an increase in the OND (the cutoff value is 2.5 mm) for brain death. The diagonal line is the nondiscriminant curve (B) Receiver operating characteristic curve showing the relationship between sensitivity and specificity to determine the predictive value of the OND (the cutoff value is 5.9 mm) for brain death. The diagonal line is the nondiscriminant line.
than age 4. An analogous study in a pediatric population was conducted by Newman et al,13 who reported an upper normal limit of 4.0 mm for patients less than 1 year old and 4.5 mm for older children. They suggested that higher OND values were strongly indicative of raised ICP. Finally, Malayeri et al4 examined children with elevated ICP observing that all patients either younger or older than age 4 exhibited OND above 4.55 mm. This study suggested that patients with brain injury who exhibited OND ⱖ 5.9 mm as well as an increased OND during sonographic monitoring of more than 2.5 mm had a high probability to progress toward brain tamponade and BD. However, the specificity and sensitivity of the above findings were rather low. Therefore, one cannot suggest that sonographic monitoring of the OND is a good candidate to predict final outcome in patients with brain injury. It is of note that alterations in the OND had a close relationship with brain CT results during the study period which may be of clinical value for the diagnosis of cerebral edema. Compared with CT, which is a highly accurate method for the evaluation of brain injury, transorbital sonography offers the significant advantages of low cost, wide availability, and rapid bedside performance. Moreover, this method is noninvasive and does not expose the patient to radiation. In this study, we used a neuroimaging scale instead of an intraventricular or subdural catheter to evaluate cerebral edema. Therefore, small variations of the ICP cannot be detected unless they sum to levels which alter the anatomical state of the brain within limits depicted by means of CT. Also, the pathophysiology of brain injury is quite complex as brain edema progresses toward brain tamponade and
therefore continuous monitoring of a variety of metabolic and circulatory parameters may be necessary.16 Transorbital sonography of the OND offers the advantage of continuous monitoring and may assist in the evaluation of brain injury, provided that examinations are performed using an accurate technique. Age and glaucoma are factors that should be taken into account when measuring the OND.31 We found an inverse relationship between age and OND which is in accord with the findings of Beatty et al.31 Also, optic nerve enlargement can occur due to secondary involvement from a variety of orbital and systematic abnormalities, such as tumor, inflammation, Grave’s disease, sarcoidosis, pseudotumor, metastasis, and hemorrhage in and around the optic nerve complex as well as hydrops from an extrinsic tumor.22 The present data showed that transorbital sonography has high intra- and interobserver reproducibility as others have also reported.17,31 However, this method relies heavily on the skill and diligence of the observer. Common technical pitfalls include: 1) inadequate depiction of the optic nerve in the transverse plane, 2) inaccurate designation of the OND contours, and 3) erroneous placement of the cursors.17 Finally, OND measurements should be performed 3 mm posterior to the papillae since the respective segment of the optic nerve features the highest level of distension and provides optimum sonographic contrast between the echogenic fat and the hypoechoic optic nerve.6,13,30 In conclusion, transorbital sonography of the OND is a highly reproducible method and may be used to evaluate brain injury and consequently cerebral edema. Alterations
3706
in the OND showed strong correlations with synchronous CT imaging findings in patients with severe brain injury. However, the sonographic monitoring of OND in the above patients showed a low specificity and sensitivity to identify patients at high risk for brain tamponade and consequently brain death. Further studies with larger numbers of patients are clearly required to determine whether sonographic monitoring of OND should be adopted more widely as a diagnostic and prognostic tool in neurocritical care. REFERENCES 1. Marshall LF, Marshall SB, Klauber MR, et al: The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 9:287, 1992 2. Barr R, Gean A: Craniofacial trauma. In Brant W, Helms C (eds): Fundamentals of Diagnostic Radiology, 2nd Ed. Philadelphia: Lippincott Williams & Wilkins; 1999, p 49 3. Czosnyka M, Pickard JD: Monitoring and interpretation of intracranial pressure. J Neurol Neurosurg Psychiatry 75:813, 2004 4. Malayeri AA, Bavarian S, Mehdizadeh M: Sonographic evaluation of optic nerve diameter in children with raised intracranial pressure. J Ultrasound Med 24:143, 2005 5. Gennarelli TA, Spielman GM, Langfitt TW, et al: Influence of the type of intracranial lesion on outcome from severe head injury. J Neurosurg 56:26, 1992 6. Hansen HC, Helmke K: The subarachnoid space surrounding the optic nerves. An ultrasound study of the optic nerve sheath. Surg Radiol Anat 18:323, 1996 7. Muller PJ, Deck JHN: Intraocular and optic nerve sheath hemorrhage in cases of sudden intracranial hypertension. J Neurosurg 41:160, 1974 8. Blaivas M, Theodoro D, Sierzenski PR: Elevated intracranial pressure detected by bedside emergency ultrasonography of optic nerve sheath. Acad Emerg Med 45:336, 2005 9. Tsung JW, Blaivas M, Cooper A, et al: A rapid noninvasive method of detecting elevated intracranial pressure using bedside ocular ultrasound: application to 3 cases of head trauma in the pediatric emergency department. Pediatr Emerg Care 21:94, 2005 10. Galetta S, Byrne SF, Smith JL: Echographic correlation of optic nerve sheath size and cerebrospinal fluid pressure. J Clin Neuroophthalmol 9:79, 1989 11. Brzezinska R, Schumacher R: Diagnosis of elevated intracranial pressure in children with shunt under special consideration of transglobe sonography of the optic nerve. Ultraschall Med 23:325, 2002 12. Helmke H, Hansen HC: Fundamentals of transorbital sonographic evaluation of optic nerve sheath expansion under intracranial hypertension. II. Patient study. Pediatr Radiol 26:706, 1996 13. Newman WD, Hollman AS, Dutton GN, et al: Measurement of optic nerve sheath diameter by ultrasound: a means of detecting acute raised intracranial pressure in hydrocephalus. Br J Ophthalmol 86:1109, 2002
KARAKITSOS, SOLDATOS, GOULIAMOS ET AL 14. Report of the medical consultants on the diagnosis of death to the President’s commission for the study of ethical problems in medicine and biomedical and behavioural research. Guidelines for the determination of death. JAMA 246:2184, 1981 15. Wijdicks EFM: The diagnosis of brain death. N Engl J Med 344:1215, 2001 16. Poularas J, Karakitsos D, Kouraklis G, et al: Comparison between transcranial colour Doppler ultrasonography and angiography in the confirmation of brain death. Transplant Proc (in press) 17. Ballantyne SA, O’Neill G, Hamilton R, et al: Observer variation in the sonographic measurement of optic nerve sheath diameter in normal adults. Eur J Ultrasound 15:145, 2002 18. Karim S, Clark RA, Poukens V, et al: Demonstration of systematic variation in human intraorbital optic nerve size by quantitative magnetic resonance imaging and histology. Invest Ophthalmol Vis Sci 45:1047, 2004 19. Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307, 1986 20. Bland JM, Altman DG: Correlation, regression, and repeated data. Br Med J 308:896, 1994 21. Sweets J: Measuring the accuracy of diagnosis system. Science 240:1285, 1988 22. Rothman MI, Zoarski GH: The Orbit. In Sutton D (ed): Textbook of Radiology and Imaging, 7th Ed. London: Churchill Livingstone; 2003, p 1573 23. Liu D, Kahn M: Measurement and relationship of subarachnoid pressure of the optic nerve to intracranial pressures in fresh cadavers. Am J Ophthalmol 116:548, 1993 24. Hansen HC, Helmke K: Validation of the optic nerve sheath response to changing cerebrospinal fluid pressure: ultrasound findings during intrathecal infusion tests. J Neurosurg 87:34, 1997 25. Killer HE, Laeng HR, Groscurth P: Lymphatic capillaries in the meninges of the human optic nerve. J Neuroophthalmol 19:222, 1999 26. Gausas RE, Gonnering RS, Lemke BN, et al: Identification of human orbital lymphatics. Ophthal Plast Reconstr Surg 15:252, 1999 27. Oehmichen M, Gruninger H, Wietholter H, et al: Lymphatic efflux of intracerebrally injected cells. Acta Neuropathol (Berl) 45:61, 1979 28. Brinker T, Ludemann W, Berens von Rautenfeld D, et al: Dynamic properties of lymphatic pathways for the absorption of cerebrospinal fluid. Acta Neuropathol (Berl) 94:493, 1997 29. Killer HE, Laeng HR, Flammer J, et al: Architecure of arachnoid trabeculae, pillars and septa in the subarachnoid space of the human optic nerve: anatomy and clinical considerations. Br J Ophthalmol 87:777, 2003 30. Helmke H, Hansen HC: Fundamentals of transorbital sonographic evaluation of optic nerve sheath expansion under intracranial hypertension. I. Experimental study. Pediatr Radiol 26:701, 1996 31. Beatty S, Good PA, McLaughlin J, et al: Echographic measurements of the retrobulbar optic nerve in normal and glaucomatous eyes. Br J Ophthalmol 82:43, 1998