- Email: [email protected]
Contrast-enhanced flair imaging in the evaluation of infectious leptomeningeal diseases
European Journal of Radiology 58 (2006) 89–95
Contrast-enhanced flair imaging in the evaluation of infectious leptomeningeal diseases夽 Hemant Parmar a,c,∗ , Yih-Yian Sitoh a , Pooja Anand b , Violet Chua a , Francis Hui a a
Department of Neuroradiology, National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore b Department of Neurology, National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore c Department of Diagnostic Imaging, The Hospital for Sick Children, Toronto, Canada Received 29 June 2005; received in revised form 20 November 2005; accepted 22 November 2005
Abstract Purpose: The purpose of our study was to compare contrast-enhanced fluid-attenuated inversion recovery (FLAIR) images with contrastenhanced T1 weighted images for infectious leptomeningitis. Materials and methods: We studied twenty-four patients with a clinical suspicion of infectious meningitis with unenhanced FLAIR, contrastenhanced T1 weighted and contrast-enhanced FLAIR MR sequences. Twelve patients had cytologic and biochemical diagnosis of meningitis on cerebrospinal fluid (CSF) examination obtained 48 h before or after the MR study. Sequences were considered positive if abnormal signal was seen in the subarachnoid space (cistern or sulci) or along pial surface. Results: Twenty-seven examinations in 24 patients were performed. Of the 12 patients (thirteen studies) in whom cytology was positive, unenhanced FLAIR images were positive in six cases (sensitivity 46%), contrast-enhanced FLAIR images were positive in 11 (sensitivity 85%), and contrast-enhanced T1 weighted MR images were positive in 11 patients (sensitivity 85%). Of the 12 patients (14 studies) in whom cerebrospinal fluid study was negative, unenhanced FLAIR images were negative in 13, contrast-enhanced FLAIR images were negative in 11, and contrast-enhanced T1 weighted MR images were negative in eight. Thus, the specificity of unenhanced FLAIR, contrast-enhanced FLAIR and contrast-enhanced T1 weighted images was 93, 79 and 57%, respectively. Conclusion: Our results suggest that post-contrast FLAIR images have similar sensitivity but a higher specificity compared to contrastenhanced T1 weighted images for detection of leptomeningeal enhancement. It can be a useful adjunct to post-contrast T1 weighted images in evaluation of infectious leptomeningitis. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: MR imaging; FLAIR; Intracranial infection; Meningitis; Leptomeningitis
1. Introduction Infectious meningitis or leptomeningitis is a serious disease with high potential for permanent neurologic impairment and high mortality rates [1]. Prompt diagnosis and early treatment is therefore, essential. While the diagnosis of 夽 This paper was presented at the 43rd Annual American Society of Neuroradiology (ASNR) meeting at Toronto, Ont., Canada, May 2005. ∗ Corresponding author at: Department of Neuroradiology, National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore. Tel.: +65 6357 7031; fax: +65 6258 1259. E-mail address: [email protected] (H. Parmar).
0720-048X/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2005.11.012
infectious meningitis is still largely based upon the clinical features and analysis of the cerebrospinal fluid (CSF), imaging-especially MR imaging, has been used with increasing frequency not only for diagnosis but also to look for any associated complication. In that respect, contrastenhanced T1 weighted imaging (WI) is a powerful and most widely used tool, particularly in distinguishing between leptomeningeal enhancement and pachymeningeal enhancement. The detection of leptomeningeal disease, however, can often be difficult because normal meninges enhance to some degree and T1 shortening due to the presence of contrast medium in normal vessels may cause confusion [2]. This effect is particularly more troublesome in children, the age
90
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
group wherein the majority of the cases of infectious meningitis are seen [1]. Fluid-attenuated inversion recovery (FLAIR) MR imaging sequence is based on an inversion recovery pulse which uses an inversion time that effectively nulls signals from the CSF. As it has a long time to echo (TE), the FLAIR sequence basically remains a T2 WI with suppression of signal from the CSF. FLAIR imaging has been reported to be a promising technique for the diagnosis of several extraaxial diseases, including subarachnoid hemorrhage and infectious and neoplastic subarachnoid space diseases [3–5]. The mild T1 effect of the FLAIR sequence responsible for contrast enhancement on these heavily T2 WI is the basis of using contrast-enhanced FLAIR images in certain brain diseases [6,7]. In a few recent studies involving limited patient numbers, post-contrast FLAIR imaging was effective in the diagnosis of parenchymal tumors and leptomeningeal diseases [2,5,6,8–10]. In their evaluation of 13 patients with meningeal carcinomatosis, Tsuchiya et al. [5] found postcontrast FLAIR images to be equivalent to conventional contrast-enhanced TI WI. While the previous papers studied leptomeningeal abnormalities due to leptomeningeal carcinomatosis or angiomatosis, we found little data on infective leptomeningitis. The purpose of our study was to compare routine FLAIR and contrast-enhanced FLAIR sequence with the conventional contrast-enhanced T1 WI to visualize infective leptomeningeal enhancement.
2. Materials and methods We studied 24 patients over a period of nine months referred for MR imaging with a clinical suspicion of infectious meningitis. MR imaging was performed on a 1.5 T clinical scanner (Signa NVi, GE Medical Systems, Milwaukee, WI, USA). All studies included an unenhanced FLAIR and contrast-enhanced T1 WI and FLAIR sequences. The timing of the contrast-enhanced T1 WI and FLAIR sequences was random; in 12 studies the FLAIR images were obtained before the contrast-enhanced T1 WI and in 15 studies it was acquired after contrast-enhanced T1 WI. Pre- and postcontrast-enhanced (Gadolinium-DTPA, 0.1 mmol/kg body weight) FLAIR sequences were obtained with flow compensation and were fast FLAIR sequences: inversion time, 2200 ms; 10,000/120 (TR/effective TE); bandwidth, 32 kHz; matrix, 256 × 192; scan time 4 min. Post-contrast T1 WI was obtained without flow compensation: 600/8 (TR/TE); bandwidth, 16 kHz; matrix, 256 × 256; scan time 3 min 25 s. In both sequences, the field of view was 24 cm, with a section thickness of 5 mm and an interslice gap of 2 mm. The unenhanced FLAIR images, contrast-enhanced FLAIR images, and contrast-enhanced T1 WI were separated from each other. Each sequence was reviewed separately by two neuroradiologists who were blinded to patient
identity, their CSF results and their clinical outcome. Each sequence was rated independently as positive or negative for leptomeningeal enhancement (Tables 2A and 2B). The sequence was considered positive when signal abnormality or enhancement was present in the subarachnoid space (cisterns and sulci) or along any pial surface or cranial nerves. The sequence was considered negative when no abnormal signal or enhancement was seen in the subarachnoid space or along the pial surfaces. The overall study was later rated into one of the three categories: (A) contrast-enhanced FLAIR better than contrast-enhanced T1 WI; (B) contrast-enhanced T1 WI better than contrast-enhanced FLAIR and (C) both the sequences similar for showing leptomeningeal enhancement. Any discordance between the two reviewers was resolved by consensus. Detailed clinical records of these patients were later obtained with particular emphasis to their CSF studies and final clinical diagnosis and outcome. In reviewing the clinical records, we considered infectious leptomeningitis to be present if positive result of CSF examination was obtained either 48 h before or after the time of MR scan.
3. Result (Table 1) A total of 27 examinations in 24 patients (three patients, patient number 1, 3 and 13-underwent imaging twice) were performed. There were 15 females and nine males. The mean patient age was 23 years (age range, 3–78 years). No supplemental oxygen was administered to any patient during their imaging studies. Twelve patients had biochemical and cytological diagnosis of infectious meningitis on CSF examination obtained 48 h before or after the MR study. Of the 12 patients (13 studies) in which CSF study was positive for leptomeningitis, the unenhanced FLAIR images were positive in six cases (sensitivity 46%), contrast-enhanced FLAIR images were positive in 11 (sensitivity 85%), and contrastenhanced T1 WI were positive in 11 patients (sensitivity 85%) (Table 2A). In three of these 13 studies, both the contrastenhanced FLAIR and T1 WI were rated as equal. Contrastenhanced FLAIR was considered better in six (Figs. 1 and 2), while T1 WI was rated better in three studies (peripheral cortical abnormalities were seen in all these three studies) (Fig. 3). In one patient (patient 17) both the contrast-enhanced T1 WI and contrast-enhanced FLAIR sequences were negative. Of the 12 patients (14 studies) in whom CSF was negative for infectious leptomeningitis, unenhanced FLAIR images were negative in 13, contrast-enhanced FLAIR images were negative in 11, and contrast-enhanced T1 weighted MR images were negative in eight. Thus, the specificity of unenhanced FLAIR, contrast-enhanced FLAIR and contrast-enhanced T1 weighted images was 93, 79 and 57%, respectively (Table 2B). There was one false positive reading among the unenhanced FLAIR images, three false positives among the contrast-enhanced FLAIR images, and six false positives among the contrast-enhanced T1 WI. There was only
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
91
Table 1 Result of individual tests in patients with suspected leptomeningeal disease Sr. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Study no.
CSF
FL/T1 (performed first)
FLAIR CM− −/+
FLAIR CM+ −/+
T1 WI CM+ −/+
Grade A/B/C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
+ + + − − − − + − + + − − − − − + + − + − + + − − + +
T1 W T1 W T1 W T1 W T1 W FL T1 W FL FL FL FL T1 W T1 W T1 W FL T1 W T1 W T1 W T1 W FL FL FL T1 W FL FL T1 W FL
+ − − − − + − + − + − − − − − − + + − − − − − − − + −
+ + + + − + + + − + + − − − − − + + − − − − + − − + +
+ + + + − + + + − − + − + + + − + + − − − + + − − + +
A A C NA NA NA NA C NA A B NA NA NA NA NA B A NA NA NA B A NA NA A C
“+”: positive,“−”: negative, NA: not applicable, Fl: FLAIR, T1 W: T1 weighted image, CM: contrast medium, CSF: cerebrospinal fuid, no.: number, A: contrast enhanced FLAIR images better than contrast-enhanced T1 WI, B: contrast-enhanced T1 WI images better than contrast-enhanced FLAIR, C: both contrast-enhanced FLAIR and contrast-enhanced T1 WI showed similar results. Table 2A MRI readings in patients with positive CSF result Imaging sequence
Positive Negative Sensitivity
FLAIR CM−
FLAIR CM+
T1 WI CM+
6 7 46%
11 2 85%
11 2 85%
CM: contrast medium,“−”: negative,“+”: positive, T1 WI: T1 weighted image.
one study (patient 19) of CSF proven leptomeningitis where contrast-enhanced T1 WI was positive compared to both the pre- and post-contrast-enhanced FLAIR sequences. Close evaluation of these FLAIR images did not reveal any pulsation artifact near the findings located by the contrastTable 2B MRI readings in patients with negative CSF result Imaging sequence
Positive Negative Sensitivity
FLAIR CM−
FLAIR CM+
T1 WI CM+
1 13 93%
3 11 79%
6 8 57%
CM: contrast medium,“−”: negative,“+”: positive, T1 WI: T1 weighted image.
enhanced T1 WI. The most common locations for observing the leptomeningeal enhancement in all sequences included the ventral surface of the brainstem (Fig. 1), the superior aspect of cerebellum, the sylvian fissures and the cerebral convexities (Fig. 2). No enhancement along any of the cranial nerve was detected in any patient on any of the three sequences studied. Diffuse dural enhancement was seen in two examinations but was not included as a positive finding. It was considered to be due to recent lumbar puncture and hence was not pathological.
4. Discussion Infectious leptomeningitis is the most common form of central nervous system (CNS) infection [11]. It can be divided into three general categories: acute pyogenic meningitis (usually bacterial infection), lymphocytic meningitis (usually viral infection) and chronic meningitis (tuberculosis and coccidiomycosis being common examples) [11]. Pathologically meningitis is characterized by breakdown of the blood–brain barrier (BBB) and vascular endothelial injury. This is followed by increased permeability across the endothelium and infiltration of polymorphonuclear cells and lymphocytes and by extensive fibrinous exudate formation [1,11]. The
92
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
Fig. 1. Twelve-year-old boy with tuberculous meningitis: (a) and (b) contrast-enhanced axial FLAIR images through the level of lower and upper brain stem show enhancing exudates coating the ventral surface of the brain stem (arrows). These exudates are less conspicuous on the contrast-enhanced axial T1 WI (c) and (d).
Fig. 2. Twenty-five-year-old man with bacterial meningitis: coronal view contrast-enhanced FLAIR image (a) show thick leptomeningeal exudates at the left occipital convexity (arrows), which is much better appreciated than on the coronal view contrast-enhanced T1 WI (b).
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
93
Fig. 3. Thirty-five-year-old lady with viral meningo-encephalitis: axial view unenhanced FLAIR (a) contrast-enhanced FLAIR (b) and contrast-enhanced T1 WI (c) images. Parenchymal abnormalities mask the leptomeningeal enhancement and sulcal hyperintensity on the contrast-enhanced FLAIR images. The enhancing exudates (arrows) were better appreciated on the Gd-T1 WI in this patient.
diagnosis of meningitis is established by the clinical history, physical examination and laboratory evaluation, especially CSF evaluation. Imaging is usually used to monitor the complications and to monitor progression. The ability to detect and differentiate different intracranial infections has, however, markedly improved with the introduction of MR imaging. The lack of bone artifact near the surface of brain parenchyma and the multiplanar capability of MR imaging have led to this pre-eminence over CT. MR imaging is also helpful in the evaluation of certain infections where CSF studies are equivocal and where detailed culture results are not available for days to weeks [12]. Herpes simplex encephalitis and tuberculous meningitis are two classic examples where MR imaging can often suggest the diagnosis before detailed CSF results (including cultures) are known.
Intravenous contrast agents are frequently used in the evaluation of BBB breakdown in various CNS disease processes such as in tumors, inflammation and infection. The most commonly used contrast agents employ gadolinium, a paramagnetic ion, which shortens both T1 and T2 relaxation times of tissues in which it has accumulated. T1 shortening is, however, the predominant effect at regular doses (0.1–0.3 mmol/kg body weight) and this effect is utilized to detect the contrast enhancement on clinical MR imaging. Hence, contrast-enhanced T1 WI are typically used for gadolinium-enhanced MR imaging [13,14] and has become the standard method of imaging nearly all intracranial infections. The detection of leptomeningeal disease, however, can often be difficult because normal meninges enhance to some degree and T1 shortening due to the presence of contrast medium in normal vessels may cause confusion [2]. Any
94
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
sequence that can reduce or eliminate signal intensity from normal vasculature constitutes a significant step forward in MR imaging [2]. FLAIR is a sensitive sequence in detecting subarachnoid space diseases and is now a part of routine MR imaging at many clinical centres. Even though FLAIR images are commonly thought of as T2-weighted images with dark cerebrospinal fluid, these images have observable T1 contrast [15], therefore, T1 shortening produced by gadolinium manifests as hyperintense signal on these images. Thus, lesions that show enhancement on post-contrast T1 WI also show enhancement on post-contrast FLAIR images [6,16,17]. The role of contrast-enhanced FLAIR in the diagnosis of infectious leptomeningitis has, however, not been established. In few previous studies, contrast-enhanced FLAIR images have been shown to be highly effective in the detection of sulcal or meningeal enhancement in CNS inflammation and metastases [2,5,6–8]. Because of the suppression of CSF signal intensity on FLAIR sequence, there is better delineation of lesions, especially meningeal lesions that abut the border of the CSF. In addition, the fact that slow-flowing blood is not usually hyperintense on post-contrast FLAIR images but frequently hyperintense on post-contrast T1 WI partly accounts for the clearer distinction between enhancing meninges and enhancing cortical veins with FLAIR imaging [8]. Our results support these findings. In six of 13 studies, the conspicuity, extension, and enhancement of the leptomeninges was superior on contrast-enhanced FLAIR images compared with post-contrast T1 WI. Contrast-enhanced T1 WI was seen to be superior to contrast-enhanced FLAIR in three out of 13 studies. In all these patients, peripheral cortical abnormalities were noted that resulted in hyperintensity on the FLAIR images and hence hindered visibility of enhancing meninges. Because contrast-enhanced FLAIR and T1 WI were acquired in a random order after contrast administration, we believe that our findings were not attributable to delayed enhancement. Deliganis had described increased signal in the CSF with increasing levels of inhaled oxygen [18]. However, in our study none of the patients had received supplemental oxygen at the time of MR imaging, negating the possibility that the high signal on FLAIR images in our patients was caused by this phenomenon. We speculate that CSF signal suppression and loss of bright signal from the blood vessels resulted in better depiction of inflamed meninges on the post-contrast FLAIR images. This may also be the reason for less false positive results on post-contrast FLAIR images compared to post-contrast T1 WI. On further review of the false positive post-contrast FLAIR images, we found the most common areas of signal abnormality at the base of the brain, where CSF pulsations are known to cause artefacts. All the false positive post-contrast T1 WI were due to excessive leptomeningeal vascular enhancement which falsely appeared to represent enhancing meninges, especially when they were seen extending for many slices. This problem was especially seen in younger patients of our group. Recently, Galassi et al. [19] have demonstrated better visualization of abnormal
meninges using fat suppressed post-gadolinium enhanced T1 weighted images, due to suppression of fat from the overlying scalp and diploic space. While this technique may improve sensitivity, its specificity has not reliably established [19,20]. We found the unenhanced FLAIR images to have low sensitivity (46%) but high specificity (93%) in detecting sulcal hyperintensity in patients with leptomeningitis. The reason for high signal on the unenhanced FLAIR images is probably related to high CSF protein content interfering with the FLAIR suppression of normal non-proteinaceous CSF. High CSF protein has T2 relaxation time longer than gray matter and will therefore, appear hyperintense with longer TE [12]. The low sensitivity of unenhanced FLAIR images, however, limits its use in the evaluation of patients with leptomeningitis. At the same time, contrast-enhanced FLAIR images become more useful if they are acquired with unenhanced FLAIR images as it helps to differentiate between the sulcal hyperintensity due to either T2 lengthening or T1 shortening. If only post-contrast FLAIR imaging is performed, areas of increased signal intensity from T2 lengthening, such as those seen with vasogenic edema, may obscure areas of BBB breakdown with resultant T1 shortening. Combination of unenhanced and contrast-enhanced FLAIR images thus leads to the best results. The use of fast FLAIR sequences with its short scan duration makes this a clinically viable tool. 5. Conclusion Our results suggest that post-contrast FLAIR images have similar sensitivity but a higher specificity compared to contrast-enhanced T1 WI for detection of subtle leptomeningeal enhancement. Contrast-enhanced FLAIR imaging, especially when combined with unenhanced FLAIR sequence, can be a valuable adjunct to post-contrast T1 WI in evaluation of patients with infective leptomeningitis, particularly when findings at gadolinium enhanced T1 WI are inconclusive. Contrast-enhanced T1 WI is more useful when associated parenchymal abnormality is present. References [1] Ashwal S, Tomasi L, Schneider S, Perkin R, Thompson J. Bacterial meningitis in children: pathophysiology and treatment. Neurology 1992;42:739–48. [2] Griffiths PD, Coley SC, Romanowski CA, Hodgson T, Wilkinson ID. Contrast-enhanced fluid-attenuated inversion recovery imaging for leptomeningeal disease in children. AJNR Am J Neuroradiol 2003;24:719–23. [3] Noguchi K, Ogawa T, Inugami A, et al. Acute subarachnoid hemorrhage: MR imaging with fluid-attenuated inversion recovery pulse sequences. Radiology 1995;196:773–7. [4] Singer MB, Atlas SW, Drayer BP. Subarachnoid space disease: diagnosis with fluid-attenuated inversion recovery MR imaging and comparison with gadolinium-enhanced spin-echo MR imaging-blinder reader study. Radiology 1998;208:417–22.
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95 [5] Tsuchiya K, Katase S, Yoshino A, Hachiya J. FLAIR MR imaging for diagnosing intracranial meningeal carcinomatosis. AJR Am J Roentgenol 2001;176:1585–8. [6] Essig M, Knopp MV, Schoenberg SO, et al. Cerebral gliomas and metastases: assessment contrast-enhanced fast-fluid attenuated inversion-recovery MR imaging. Radiology 1999;210:551–7. [7] Ercan N, Gultekin S, Celik H, Tali TE, Oner YA, Erbas G. Diagnostic value of contrast-enhanced fluid-attenuated inversion recovery MR imaging of intracranial metastases. AJNR Am J Neuroradiol 2004;25:761–5. [8] Mathews VP, Caldemeyer KS, Lowe MJ, Greenspan SL, Weber DM, Ulmer JL. Brain: gadolinium-enhanced fast fluid-attenuated inversion-recovery MR imaging. Radiology 1999;211:257–63. [9] Essig M, Schlemmer HP, Tronnier V, Hawighorst H, Wirtz R, van Kaick G. Fluid-attenuated inversion-recovery MR imaging of gliomatosis cerebri. Eur Radiol 2001;11:303–8. [10] Essig M, Metzner R, Bonsanto M, et al. Postoperative fluidattenuated inversion-recovery MR imaging of cerebral gliomas: initial results. Eur Radiol 2001;11:2004–10. [11] Osborn AG. Infections of the brain and its linings. Diagnostic neuroradiology. St. Louis, CV: Mosby; 1994. p. 680–2. [12] Kamran S, Bener AB, Alper D, Bakshi R. Role of fluid-attenuated inversion recovery in the diagnosis of meningitis. JCAT J Comput Assist Tomogr 2004;28:68–72. [13] Mathews VP, Caldemeyer KS, Ulmer JL, Nguyen H, Yuh WT. Effects of contrast dose, delayed imaging and magnetization trans-
[14]
[15]
[16]
[17]
[18]
[19]
[20]
95
fer on gadolinium-enhanced MR imaging of brain lesions. JMRI 1997;7:14–22. Nelson KL, Runge VM. Principles of MR contrast. In: Runge VM, editor. Contrast-enhanced clinical magnetic resonance imaging. Lexington, KY: University of Kentucky Press; 1997. p. 1–13. Hendrick RE, Raff U. Image contrast and noise. In: Stark DD, Bradley WG, editors. Magnetic resonance imaging. 2nd ed. St. Louis, MO: Mosby-Yearbook; 1992. p. 123–9. Tsuchiya K, Katase S, Yoshino A, et al. Use of contrast enhanced fluid-attenuated inversion recovery magnetic resonance imaging in the diagnosis of metastatic brain tumors. Int J Neuroradiol 1999;5:9–15. Melhem ER, Bert RJ, Walker RE. Usefulness of optimized gadolinium-enhanced fast fluid-attenuated inversion recovery MR imaging in revealing lesions of the brain. AJR Am J Roentgenol 1998;171:803–7. Deliganis AV, Fisher DJ, Lam AM, Maravilla KR. CSF signal intensity increase on FLAIR MR images in patients under general anesthesia: the role of supplemental oxygen. Radiology 2001;218:152–6. Galassi W, Phutthrak W, Hesselink JR, Healy JF, Dietrich RB, Imbesi SG. Intracranial meningeal disease: comparison of contrastenhanced MR imaging with fluid attenuated inversion recovery and fat-suppressed T1-weighted sequences. AJNR Am J Neuroradiol 2005;26:553–9. Mathews VP. Show me the gadolinium. AJNR Am J Neuroradiol 2005;26:440–1.
Contrast-enhanced flair imaging in the evaluation of infectious leptomeningeal diseases夽 Hemant Parmar a,c,∗ , Yih-Yian Sitoh a , Pooja Anand b , Violet Chua a , Francis Hui a a
Department of Neuroradiology, National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore b Department of Neurology, National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore c Department of Diagnostic Imaging, The Hospital for Sick Children, Toronto, Canada Received 29 June 2005; received in revised form 20 November 2005; accepted 22 November 2005
Abstract Purpose: The purpose of our study was to compare contrast-enhanced fluid-attenuated inversion recovery (FLAIR) images with contrastenhanced T1 weighted images for infectious leptomeningitis. Materials and methods: We studied twenty-four patients with a clinical suspicion of infectious meningitis with unenhanced FLAIR, contrastenhanced T1 weighted and contrast-enhanced FLAIR MR sequences. Twelve patients had cytologic and biochemical diagnosis of meningitis on cerebrospinal fluid (CSF) examination obtained 48 h before or after the MR study. Sequences were considered positive if abnormal signal was seen in the subarachnoid space (cistern or sulci) or along pial surface. Results: Twenty-seven examinations in 24 patients were performed. Of the 12 patients (thirteen studies) in whom cytology was positive, unenhanced FLAIR images were positive in six cases (sensitivity 46%), contrast-enhanced FLAIR images were positive in 11 (sensitivity 85%), and contrast-enhanced T1 weighted MR images were positive in 11 patients (sensitivity 85%). Of the 12 patients (14 studies) in whom cerebrospinal fluid study was negative, unenhanced FLAIR images were negative in 13, contrast-enhanced FLAIR images were negative in 11, and contrast-enhanced T1 weighted MR images were negative in eight. Thus, the specificity of unenhanced FLAIR, contrast-enhanced FLAIR and contrast-enhanced T1 weighted images was 93, 79 and 57%, respectively. Conclusion: Our results suggest that post-contrast FLAIR images have similar sensitivity but a higher specificity compared to contrastenhanced T1 weighted images for detection of leptomeningeal enhancement. It can be a useful adjunct to post-contrast T1 weighted images in evaluation of infectious leptomeningitis. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: MR imaging; FLAIR; Intracranial infection; Meningitis; Leptomeningitis
1. Introduction Infectious meningitis or leptomeningitis is a serious disease with high potential for permanent neurologic impairment and high mortality rates [1]. Prompt diagnosis and early treatment is therefore, essential. While the diagnosis of 夽 This paper was presented at the 43rd Annual American Society of Neuroradiology (ASNR) meeting at Toronto, Ont., Canada, May 2005. ∗ Corresponding author at: Department of Neuroradiology, National Neuroscience Institute, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore. Tel.: +65 6357 7031; fax: +65 6258 1259. E-mail address: [email protected] (H. Parmar).
0720-048X/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2005.11.012
infectious meningitis is still largely based upon the clinical features and analysis of the cerebrospinal fluid (CSF), imaging-especially MR imaging, has been used with increasing frequency not only for diagnosis but also to look for any associated complication. In that respect, contrastenhanced T1 weighted imaging (WI) is a powerful and most widely used tool, particularly in distinguishing between leptomeningeal enhancement and pachymeningeal enhancement. The detection of leptomeningeal disease, however, can often be difficult because normal meninges enhance to some degree and T1 shortening due to the presence of contrast medium in normal vessels may cause confusion [2]. This effect is particularly more troublesome in children, the age
90
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
group wherein the majority of the cases of infectious meningitis are seen [1]. Fluid-attenuated inversion recovery (FLAIR) MR imaging sequence is based on an inversion recovery pulse which uses an inversion time that effectively nulls signals from the CSF. As it has a long time to echo (TE), the FLAIR sequence basically remains a T2 WI with suppression of signal from the CSF. FLAIR imaging has been reported to be a promising technique for the diagnosis of several extraaxial diseases, including subarachnoid hemorrhage and infectious and neoplastic subarachnoid space diseases [3–5]. The mild T1 effect of the FLAIR sequence responsible for contrast enhancement on these heavily T2 WI is the basis of using contrast-enhanced FLAIR images in certain brain diseases [6,7]. In a few recent studies involving limited patient numbers, post-contrast FLAIR imaging was effective in the diagnosis of parenchymal tumors and leptomeningeal diseases [2,5,6,8–10]. In their evaluation of 13 patients with meningeal carcinomatosis, Tsuchiya et al. [5] found postcontrast FLAIR images to be equivalent to conventional contrast-enhanced TI WI. While the previous papers studied leptomeningeal abnormalities due to leptomeningeal carcinomatosis or angiomatosis, we found little data on infective leptomeningitis. The purpose of our study was to compare routine FLAIR and contrast-enhanced FLAIR sequence with the conventional contrast-enhanced T1 WI to visualize infective leptomeningeal enhancement.
2. Materials and methods We studied 24 patients over a period of nine months referred for MR imaging with a clinical suspicion of infectious meningitis. MR imaging was performed on a 1.5 T clinical scanner (Signa NVi, GE Medical Systems, Milwaukee, WI, USA). All studies included an unenhanced FLAIR and contrast-enhanced T1 WI and FLAIR sequences. The timing of the contrast-enhanced T1 WI and FLAIR sequences was random; in 12 studies the FLAIR images were obtained before the contrast-enhanced T1 WI and in 15 studies it was acquired after contrast-enhanced T1 WI. Pre- and postcontrast-enhanced (Gadolinium-DTPA, 0.1 mmol/kg body weight) FLAIR sequences were obtained with flow compensation and were fast FLAIR sequences: inversion time, 2200 ms; 10,000/120 (TR/effective TE); bandwidth, 32 kHz; matrix, 256 × 192; scan time 4 min. Post-contrast T1 WI was obtained without flow compensation: 600/8 (TR/TE); bandwidth, 16 kHz; matrix, 256 × 256; scan time 3 min 25 s. In both sequences, the field of view was 24 cm, with a section thickness of 5 mm and an interslice gap of 2 mm. The unenhanced FLAIR images, contrast-enhanced FLAIR images, and contrast-enhanced T1 WI were separated from each other. Each sequence was reviewed separately by two neuroradiologists who were blinded to patient
identity, their CSF results and their clinical outcome. Each sequence was rated independently as positive or negative for leptomeningeal enhancement (Tables 2A and 2B). The sequence was considered positive when signal abnormality or enhancement was present in the subarachnoid space (cisterns and sulci) or along any pial surface or cranial nerves. The sequence was considered negative when no abnormal signal or enhancement was seen in the subarachnoid space or along the pial surfaces. The overall study was later rated into one of the three categories: (A) contrast-enhanced FLAIR better than contrast-enhanced T1 WI; (B) contrast-enhanced T1 WI better than contrast-enhanced FLAIR and (C) both the sequences similar for showing leptomeningeal enhancement. Any discordance between the two reviewers was resolved by consensus. Detailed clinical records of these patients were later obtained with particular emphasis to their CSF studies and final clinical diagnosis and outcome. In reviewing the clinical records, we considered infectious leptomeningitis to be present if positive result of CSF examination was obtained either 48 h before or after the time of MR scan.
3. Result (Table 1) A total of 27 examinations in 24 patients (three patients, patient number 1, 3 and 13-underwent imaging twice) were performed. There were 15 females and nine males. The mean patient age was 23 years (age range, 3–78 years). No supplemental oxygen was administered to any patient during their imaging studies. Twelve patients had biochemical and cytological diagnosis of infectious meningitis on CSF examination obtained 48 h before or after the MR study. Of the 12 patients (13 studies) in which CSF study was positive for leptomeningitis, the unenhanced FLAIR images were positive in six cases (sensitivity 46%), contrast-enhanced FLAIR images were positive in 11 (sensitivity 85%), and contrastenhanced T1 WI were positive in 11 patients (sensitivity 85%) (Table 2A). In three of these 13 studies, both the contrastenhanced FLAIR and T1 WI were rated as equal. Contrastenhanced FLAIR was considered better in six (Figs. 1 and 2), while T1 WI was rated better in three studies (peripheral cortical abnormalities were seen in all these three studies) (Fig. 3). In one patient (patient 17) both the contrast-enhanced T1 WI and contrast-enhanced FLAIR sequences were negative. Of the 12 patients (14 studies) in whom CSF was negative for infectious leptomeningitis, unenhanced FLAIR images were negative in 13, contrast-enhanced FLAIR images were negative in 11, and contrast-enhanced T1 weighted MR images were negative in eight. Thus, the specificity of unenhanced FLAIR, contrast-enhanced FLAIR and contrast-enhanced T1 weighted images was 93, 79 and 57%, respectively (Table 2B). There was one false positive reading among the unenhanced FLAIR images, three false positives among the contrast-enhanced FLAIR images, and six false positives among the contrast-enhanced T1 WI. There was only
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
91
Table 1 Result of individual tests in patients with suspected leptomeningeal disease Sr. no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Study no.
CSF
FL/T1 (performed first)
FLAIR CM− −/+
FLAIR CM+ −/+
T1 WI CM+ −/+
Grade A/B/C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
+ + + − − − − + − + + − − − − − + + − + − + + − − + +
T1 W T1 W T1 W T1 W T1 W FL T1 W FL FL FL FL T1 W T1 W T1 W FL T1 W T1 W T1 W T1 W FL FL FL T1 W FL FL T1 W FL
+ − − − − + − + − + − − − − − − + + − − − − − − − + −
+ + + + − + + + − + + − − − − − + + − − − − + − − + +
+ + + + − + + + − − + − + + + − + + − − − + + − − + +
A A C NA NA NA NA C NA A B NA NA NA NA NA B A NA NA NA B A NA NA A C
“+”: positive,“−”: negative, NA: not applicable, Fl: FLAIR, T1 W: T1 weighted image, CM: contrast medium, CSF: cerebrospinal fuid, no.: number, A: contrast enhanced FLAIR images better than contrast-enhanced T1 WI, B: contrast-enhanced T1 WI images better than contrast-enhanced FLAIR, C: both contrast-enhanced FLAIR and contrast-enhanced T1 WI showed similar results. Table 2A MRI readings in patients with positive CSF result Imaging sequence
Positive Negative Sensitivity
FLAIR CM−
FLAIR CM+
T1 WI CM+
6 7 46%
11 2 85%
11 2 85%
CM: contrast medium,“−”: negative,“+”: positive, T1 WI: T1 weighted image.
one study (patient 19) of CSF proven leptomeningitis where contrast-enhanced T1 WI was positive compared to both the pre- and post-contrast-enhanced FLAIR sequences. Close evaluation of these FLAIR images did not reveal any pulsation artifact near the findings located by the contrastTable 2B MRI readings in patients with negative CSF result Imaging sequence
Positive Negative Sensitivity
FLAIR CM−
FLAIR CM+
T1 WI CM+
1 13 93%
3 11 79%
6 8 57%
CM: contrast medium,“−”: negative,“+”: positive, T1 WI: T1 weighted image.
enhanced T1 WI. The most common locations for observing the leptomeningeal enhancement in all sequences included the ventral surface of the brainstem (Fig. 1), the superior aspect of cerebellum, the sylvian fissures and the cerebral convexities (Fig. 2). No enhancement along any of the cranial nerve was detected in any patient on any of the three sequences studied. Diffuse dural enhancement was seen in two examinations but was not included as a positive finding. It was considered to be due to recent lumbar puncture and hence was not pathological.
4. Discussion Infectious leptomeningitis is the most common form of central nervous system (CNS) infection [11]. It can be divided into three general categories: acute pyogenic meningitis (usually bacterial infection), lymphocytic meningitis (usually viral infection) and chronic meningitis (tuberculosis and coccidiomycosis being common examples) [11]. Pathologically meningitis is characterized by breakdown of the blood–brain barrier (BBB) and vascular endothelial injury. This is followed by increased permeability across the endothelium and infiltration of polymorphonuclear cells and lymphocytes and by extensive fibrinous exudate formation [1,11]. The
92
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
Fig. 1. Twelve-year-old boy with tuberculous meningitis: (a) and (b) contrast-enhanced axial FLAIR images through the level of lower and upper brain stem show enhancing exudates coating the ventral surface of the brain stem (arrows). These exudates are less conspicuous on the contrast-enhanced axial T1 WI (c) and (d).
Fig. 2. Twenty-five-year-old man with bacterial meningitis: coronal view contrast-enhanced FLAIR image (a) show thick leptomeningeal exudates at the left occipital convexity (arrows), which is much better appreciated than on the coronal view contrast-enhanced T1 WI (b).
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
93
Fig. 3. Thirty-five-year-old lady with viral meningo-encephalitis: axial view unenhanced FLAIR (a) contrast-enhanced FLAIR (b) and contrast-enhanced T1 WI (c) images. Parenchymal abnormalities mask the leptomeningeal enhancement and sulcal hyperintensity on the contrast-enhanced FLAIR images. The enhancing exudates (arrows) were better appreciated on the Gd-T1 WI in this patient.
diagnosis of meningitis is established by the clinical history, physical examination and laboratory evaluation, especially CSF evaluation. Imaging is usually used to monitor the complications and to monitor progression. The ability to detect and differentiate different intracranial infections has, however, markedly improved with the introduction of MR imaging. The lack of bone artifact near the surface of brain parenchyma and the multiplanar capability of MR imaging have led to this pre-eminence over CT. MR imaging is also helpful in the evaluation of certain infections where CSF studies are equivocal and where detailed culture results are not available for days to weeks [12]. Herpes simplex encephalitis and tuberculous meningitis are two classic examples where MR imaging can often suggest the diagnosis before detailed CSF results (including cultures) are known.
Intravenous contrast agents are frequently used in the evaluation of BBB breakdown in various CNS disease processes such as in tumors, inflammation and infection. The most commonly used contrast agents employ gadolinium, a paramagnetic ion, which shortens both T1 and T2 relaxation times of tissues in which it has accumulated. T1 shortening is, however, the predominant effect at regular doses (0.1–0.3 mmol/kg body weight) and this effect is utilized to detect the contrast enhancement on clinical MR imaging. Hence, contrast-enhanced T1 WI are typically used for gadolinium-enhanced MR imaging [13,14] and has become the standard method of imaging nearly all intracranial infections. The detection of leptomeningeal disease, however, can often be difficult because normal meninges enhance to some degree and T1 shortening due to the presence of contrast medium in normal vessels may cause confusion [2]. Any
94
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95
sequence that can reduce or eliminate signal intensity from normal vasculature constitutes a significant step forward in MR imaging [2]. FLAIR is a sensitive sequence in detecting subarachnoid space diseases and is now a part of routine MR imaging at many clinical centres. Even though FLAIR images are commonly thought of as T2-weighted images with dark cerebrospinal fluid, these images have observable T1 contrast [15], therefore, T1 shortening produced by gadolinium manifests as hyperintense signal on these images. Thus, lesions that show enhancement on post-contrast T1 WI also show enhancement on post-contrast FLAIR images [6,16,17]. The role of contrast-enhanced FLAIR in the diagnosis of infectious leptomeningitis has, however, not been established. In few previous studies, contrast-enhanced FLAIR images have been shown to be highly effective in the detection of sulcal or meningeal enhancement in CNS inflammation and metastases [2,5,6–8]. Because of the suppression of CSF signal intensity on FLAIR sequence, there is better delineation of lesions, especially meningeal lesions that abut the border of the CSF. In addition, the fact that slow-flowing blood is not usually hyperintense on post-contrast FLAIR images but frequently hyperintense on post-contrast T1 WI partly accounts for the clearer distinction between enhancing meninges and enhancing cortical veins with FLAIR imaging [8]. Our results support these findings. In six of 13 studies, the conspicuity, extension, and enhancement of the leptomeninges was superior on contrast-enhanced FLAIR images compared with post-contrast T1 WI. Contrast-enhanced T1 WI was seen to be superior to contrast-enhanced FLAIR in three out of 13 studies. In all these patients, peripheral cortical abnormalities were noted that resulted in hyperintensity on the FLAIR images and hence hindered visibility of enhancing meninges. Because contrast-enhanced FLAIR and T1 WI were acquired in a random order after contrast administration, we believe that our findings were not attributable to delayed enhancement. Deliganis had described increased signal in the CSF with increasing levels of inhaled oxygen [18]. However, in our study none of the patients had received supplemental oxygen at the time of MR imaging, negating the possibility that the high signal on FLAIR images in our patients was caused by this phenomenon. We speculate that CSF signal suppression and loss of bright signal from the blood vessels resulted in better depiction of inflamed meninges on the post-contrast FLAIR images. This may also be the reason for less false positive results on post-contrast FLAIR images compared to post-contrast T1 WI. On further review of the false positive post-contrast FLAIR images, we found the most common areas of signal abnormality at the base of the brain, where CSF pulsations are known to cause artefacts. All the false positive post-contrast T1 WI were due to excessive leptomeningeal vascular enhancement which falsely appeared to represent enhancing meninges, especially when they were seen extending for many slices. This problem was especially seen in younger patients of our group. Recently, Galassi et al. [19] have demonstrated better visualization of abnormal
meninges using fat suppressed post-gadolinium enhanced T1 weighted images, due to suppression of fat from the overlying scalp and diploic space. While this technique may improve sensitivity, its specificity has not reliably established [19,20]. We found the unenhanced FLAIR images to have low sensitivity (46%) but high specificity (93%) in detecting sulcal hyperintensity in patients with leptomeningitis. The reason for high signal on the unenhanced FLAIR images is probably related to high CSF protein content interfering with the FLAIR suppression of normal non-proteinaceous CSF. High CSF protein has T2 relaxation time longer than gray matter and will therefore, appear hyperintense with longer TE [12]. The low sensitivity of unenhanced FLAIR images, however, limits its use in the evaluation of patients with leptomeningitis. At the same time, contrast-enhanced FLAIR images become more useful if they are acquired with unenhanced FLAIR images as it helps to differentiate between the sulcal hyperintensity due to either T2 lengthening or T1 shortening. If only post-contrast FLAIR imaging is performed, areas of increased signal intensity from T2 lengthening, such as those seen with vasogenic edema, may obscure areas of BBB breakdown with resultant T1 shortening. Combination of unenhanced and contrast-enhanced FLAIR images thus leads to the best results. The use of fast FLAIR sequences with its short scan duration makes this a clinically viable tool. 5. Conclusion Our results suggest that post-contrast FLAIR images have similar sensitivity but a higher specificity compared to contrast-enhanced T1 WI for detection of subtle leptomeningeal enhancement. Contrast-enhanced FLAIR imaging, especially when combined with unenhanced FLAIR sequence, can be a valuable adjunct to post-contrast T1 WI in evaluation of patients with infective leptomeningitis, particularly when findings at gadolinium enhanced T1 WI are inconclusive. Contrast-enhanced T1 WI is more useful when associated parenchymal abnormality is present. References [1] Ashwal S, Tomasi L, Schneider S, Perkin R, Thompson J. Bacterial meningitis in children: pathophysiology and treatment. Neurology 1992;42:739–48. [2] Griffiths PD, Coley SC, Romanowski CA, Hodgson T, Wilkinson ID. Contrast-enhanced fluid-attenuated inversion recovery imaging for leptomeningeal disease in children. AJNR Am J Neuroradiol 2003;24:719–23. [3] Noguchi K, Ogawa T, Inugami A, et al. Acute subarachnoid hemorrhage: MR imaging with fluid-attenuated inversion recovery pulse sequences. Radiology 1995;196:773–7. [4] Singer MB, Atlas SW, Drayer BP. Subarachnoid space disease: diagnosis with fluid-attenuated inversion recovery MR imaging and comparison with gadolinium-enhanced spin-echo MR imaging-blinder reader study. Radiology 1998;208:417–22.
H. Parmar et al. / European Journal of Radiology 58 (2006) 89–95 [5] Tsuchiya K, Katase S, Yoshino A, Hachiya J. FLAIR MR imaging for diagnosing intracranial meningeal carcinomatosis. AJR Am J Roentgenol 2001;176:1585–8. [6] Essig M, Knopp MV, Schoenberg SO, et al. Cerebral gliomas and metastases: assessment contrast-enhanced fast-fluid attenuated inversion-recovery MR imaging. Radiology 1999;210:551–7. [7] Ercan N, Gultekin S, Celik H, Tali TE, Oner YA, Erbas G. Diagnostic value of contrast-enhanced fluid-attenuated inversion recovery MR imaging of intracranial metastases. AJNR Am J Neuroradiol 2004;25:761–5. [8] Mathews VP, Caldemeyer KS, Lowe MJ, Greenspan SL, Weber DM, Ulmer JL. Brain: gadolinium-enhanced fast fluid-attenuated inversion-recovery MR imaging. Radiology 1999;211:257–63. [9] Essig M, Schlemmer HP, Tronnier V, Hawighorst H, Wirtz R, van Kaick G. Fluid-attenuated inversion-recovery MR imaging of gliomatosis cerebri. Eur Radiol 2001;11:303–8. [10] Essig M, Metzner R, Bonsanto M, et al. Postoperative fluidattenuated inversion-recovery MR imaging of cerebral gliomas: initial results. Eur Radiol 2001;11:2004–10. [11] Osborn AG. Infections of the brain and its linings. Diagnostic neuroradiology. St. Louis, CV: Mosby; 1994. p. 680–2. [12] Kamran S, Bener AB, Alper D, Bakshi R. Role of fluid-attenuated inversion recovery in the diagnosis of meningitis. JCAT J Comput Assist Tomogr 2004;28:68–72. [13] Mathews VP, Caldemeyer KS, Ulmer JL, Nguyen H, Yuh WT. Effects of contrast dose, delayed imaging and magnetization trans-
[14]
[15]
[16]
[17]
[18]
[19]
[20]
95
fer on gadolinium-enhanced MR imaging of brain lesions. JMRI 1997;7:14–22. Nelson KL, Runge VM. Principles of MR contrast. In: Runge VM, editor. Contrast-enhanced clinical magnetic resonance imaging. Lexington, KY: University of Kentucky Press; 1997. p. 1–13. Hendrick RE, Raff U. Image contrast and noise. In: Stark DD, Bradley WG, editors. Magnetic resonance imaging. 2nd ed. St. Louis, MO: Mosby-Yearbook; 1992. p. 123–9. Tsuchiya K, Katase S, Yoshino A, et al. Use of contrast enhanced fluid-attenuated inversion recovery magnetic resonance imaging in the diagnosis of metastatic brain tumors. Int J Neuroradiol 1999;5:9–15. Melhem ER, Bert RJ, Walker RE. Usefulness of optimized gadolinium-enhanced fast fluid-attenuated inversion recovery MR imaging in revealing lesions of the brain. AJR Am J Roentgenol 1998;171:803–7. Deliganis AV, Fisher DJ, Lam AM, Maravilla KR. CSF signal intensity increase on FLAIR MR images in patients under general anesthesia: the role of supplemental oxygen. Radiology 2001;218:152–6. Galassi W, Phutthrak W, Hesselink JR, Healy JF, Dietrich RB, Imbesi SG. Intracranial meningeal disease: comparison of contrastenhanced MR imaging with fluid attenuated inversion recovery and fat-suppressed T1-weighted sequences. AJNR Am J Neuroradiol 2005;26:553–9. Mathews VP. Show me the gadolinium. AJNR Am J Neuroradiol 2005;26:440–1.