Prospective evaluation of in vivo proton MR spectroscopy in differentiation of similar appearing intracranial cystic lesions

Transcription

1 Magnetic Resonance Imaging 19 (2001) Prospective evaluation of in vivo proton MR spectroscopy in differentiation of similar appearing intracranial cystic lesions A. Shukla-Dave a, R. K. Gupta a, *, R. Roy d, N. Husain e, L. Paul b, S. K. Venkatesh a, M. R. Rashid f, D. K. Chhabra c, M. Husain f a Department of Radiology, Sanjay Gandhi Post-Graduate Institute of Medical Sciences, Lucknow , India b Department of Pathology, Sanjay Gandhi Post-Graduate Institute of Medical Sciences, Lucknow , India c Department of Neurosurgery, Sanjay Gandhi Post-Graduate Institute of Medical Sciences, Lucknow , India d Regional Sophisticated Instrumentation Center, CDRI, Lucknow, India e Department of Pathology, KG Medical College, Lucknow, India f Department of Neurosurgery, KG Medical College, Lucknow, India Received 2 September 2000; accepted 20 December 2000 Abstract Proton MR spectroscopy (PMRS) has been found to be useful in differentiating various cystic intracranial lesions. The purpose of the present study was to prospectively evaluate the spectral pattern of various cystic lesions of brain with similar imaging appearances and to determine the accuracy of this technique in the differential diagnosis of these lesions. Fifty-one patients with intracranial cystic lesions (21 abscesses, 20 gliomas, 3 hydatid cysts, 3 arachnoid cysts, 1 case each of glioependymal cyst, xanthogranuloma, infarction and acoustic neuroma) were evaluated with conventional MR imaging and in vivo PMRS. Ex vivo PMRS of the cystic contents aspirated at surgery in 31 cases was also done to confirm the in-vivo results. Preoperative diagnosis of the lesions was based on the results of in vivo PMRS. In vivo PMRS accurately predicted the pathology in 92% of the cases. We conclude that in-vivo PMRS complements imaging in better characterization of cystic intracranial mass lesions Elsevier Science Inc. All rights reserved. Keywords: MR Spectroscopy; Cystic Lesions; Brain; MRI 1. Introduction Cystic intracranial lesions include true cysts lined by epithelial, ependymal, or meningothelial cells, dermoid and epidermoid cysts, parasitic cysts (cysticercosis, hydatid cysts) or may be pseudocystic neoplastic or inflammatory lesions secondary to accumulation of necrotic, intercellular mixed or proteinaceous material [1 3]. It is not always possible to differentiate different cystic intracranial lesions on MR imaging findings. The management of these cystic lesions depends upon its nature and it varies from definite surgery (cystic glioma) to minimal invasion (abscesses, arachnoid cyst) [4,5]. Newer imaging techniques such as diffusion weighted imaging (DWI) and fluid attenuated inversion recovery imaging (FLAIR) have been applied to differentiate the cystic lesions [4 6]. DWI is considered * Corresponding author. Tel.: ; fax: , address: rgupta@sgpgi.ac.in (R. Gupta). useful in differentiating cystic gliomas from cerebral abscesses [5,6] and there are no reports of its use in differentiation in other types of cystic lesions. FLAIR imaging has also been used in the differentiation of cystic intracranial lesions. It has been shown that the maldevelopmental/porencephalic cysts can be differentiated from neoplastic/inflammatory lesions [4]. However, FLAIR sequence was not found to be useful in separating neoplastic lesions from inflammatory/infective cystic lesions [4]. In vivo proton MR spectroscopy (MRS) has shown a high potential in the differentiation of neoplastic from non-neoplastic tissue [3,6 12]. Most of these earlier studies are of retrospective nature. The purpose of present study was to evaluate in prospective the accuracy of PMRS in differentiation of intracranial cystic lesions. 2. Patients and methods Fifty-one patients (23 men and 28 women; mean age 32.6 years; range 8 to 50 years) having similar appearing X/01/$ see front matter 2001 Elsevier Science Inc. All rights reserved. PII: S X(01)

2 104 A. Shukla-Dave et al. / Magnetic Resonance Imaging 19 (2001) Table 1 Summary of results No. of cases MR Spectroscopic findings MRS Diagnosis Final diagnosis 21 Ala, Ac, Suc, Lac, CytoAA Abscess Abscess 19 Cho, Lip/Lac Glioma Glioma 3 Lac Arachnoid cyst Arachnoid cyst 3 Lac, Ac, Suc, Hydatid cyst Hydatid cyst 2 Cho, Lip/Lac Glioma Xanthogranuloma (1), Infarct (1) 1 Lac Inconclusive Glioblastoma multiforme 1 *Single resonance at 2.04 ppm Inconclusive Glioependymal cyst 1 Poor quality data Inconclusive Acoustic neuroma Cho Choline (3/2 ppm); Lip/Lac Lipid/Lactate (1.3 ppm); Ala Alanine (1.5 ppm); Ac acetate (1.92 ppm); Sue Succinate (2.4 ppm); CytoAA cytosolic amino acids (Leucine, Isoleucine, Valine) (0.9 ppm); * assigned to N-acetyl group of compound. intracranial cystic lesions on conventional MR imaging formed the study group. The cystic lesions included 21 abscesses, 20 gliomas, 3 arachnoid cysts, 3 hydatid cysts and 1 case each of glioependymal cyst, xanthogranuloma, infarction and acoustic neuroma. All the patients were subjected to conventional MR imaging including proton density (PD), T2 and T1 weighted sequences. In-vivo PMRS was done with a single voxel technique and was interpreted independently by two investigators (RKG, ASD) with out the knowledge of imaging except that the lesions were purely cystic. Ex vivo PMRS was interpreted independently by another author (RR) who was blinded of the imaging and in vivo PMRS data. The pre-operative diagnosis was based entirely on PMRS findings. Ex-vivo MRS of the fluid contents of cysts aspirated at surgery was also carried out in 31 patients. The final diagnosis was based on the results of histopathology, aspiration and culture of the contents for the organism (abscesses). The results of PMRS diagnosis were compared with final diagnosis to know the utility of in vivo PMRS in these cystic intracranial lesions. Spectral data were available from 50 age and sex matched healthy volunteers from the control data bank. MRI and MRS Techniques: All patients underwent MRI and PMRS in one session. MRI and single voxel PMRS was performed on a 2 Tesla whole body system (Magnetom, Siemens) operating at field strength of 1.5 Tesla using a circularly polarized head coil. T 1 -weighted (TR/TE 600/15) and PD, T2 weighted, (TR/TE 2200/12/80) axial images with interslice gap of 0.5 mm, matrix were obtained. Post contrast T1 weighted axial imaging was done after injecting 0.1 mmol/kg body weight of Gadolinium-DTPA. In vivo PMRS was performed on a lesion with a size more than 8 cm 3. A volume of interest (VOI) of 4 8 ml within the confines of the lesion was selected for in-vivo MR spectroscopy. In some cases, the rim of the cystic lesion was also included in the VOI. Volume-selective spectroscopy was performed using STEAM localizing sequence with TE 20 msec, SE sequence with TE 135 msec or both in all the cases. The total time taken for imaging and spectroscopy ranged between min. Ex-vivo MRS: High resolution NMR spectroscopy of the aspirate fluid of the cystic lesions was performed in 31 cases. A 300 MHz NMR system (Bruker, Switzerland) with a multinuclear probe head and Z gradient was used with the following parameters: flip angle 90, relaxation delay 3 sec, spectral width KHz, transients 128 with presaturation of water. In addition, SE Fourier transformed spectra were also recorded with TE 80 msec and relaxation delay of 2 s to see the phase reversal of the J coupled multiplets. The diagnosis of different intracranial cystic lesions was based on the combination of resonances of different metabolites, described in the literature [3,8,13]. The criteria for the diagnosis in these cystic lesions were: 3. Results Abscesses: Lipid/lactate at 1.3 and amino acids at 0.9 ppm in all with/without additional resonances of succinate, acetate, alanine and glycine. Glioma: lipid and or lactate with choline Arachnoid cyst: Presence of small resonance of lactate with very low signal to noise spectrum Hydatid cyst: Very large succinate with lactate, alanine, acetate with absence of amino acids All the cystic lesions appeared as hyperintense on T2 weighted images and hypointense on T1 weighted images with a well-defined rim. Spectral quality was interpretable in 50 cases. In one case of acoustic neuroma, data set was of poor quality and was not included for analysis. The pathology was correctly recognized with PMRS data in 46/50 patients. The results are summarized in Table 1. In all cases of glioma (n 20) except one, choline (3.2ppm) and lipid/lactate resonances (1.3ppm) were observed. In remaining one case (Fig. 1), only lactate resonance was observed and the PMRS data were interpreted as inconclusive. In this patient, the ex vivo PMRS also did not show any choline containing compound. Resonances from cytosolic amino acids (0.9ppm) and lactate (1.3ppm) were

3 A. Shukla-Dave et al. / Magnetic Resonance Imaging 19 (2001) Fig. 1. Glioblastoma multiforme. T2 W axial image (a) shows evidence of a cystic area in the left frontal region with hypointense rim and perifocal edema. In vivo PMRS done from the inset in (a) with SE 135 ms (b) shows only lactate (L) at 1.3ppm. EX-vivo PMRS (c and d) confirms the presence of lipid and lactate (Lip & Lac) with no evidence of choline. Histopathology (e) shows marked nuclear pleomorphism, giant cells and mitotic figures [H & E X 200]. observed in all 21 cases of abscesses. Alanine (1.5ppm) (n 11), acetate (1.9ppm) (n 10), and succinate (2.4ppm) (n 7) were also observed in some abscesses. Three cases of hydatid cysts demonstrated resonances from lactate, acetate and succinate. In one patient with a hydatid cyst (Fig. 2) with perifocal edema, choline and resonances at 3.3 and 3.4 ppm were also observed at STEAM 20 msec. The resonances at 3.3 ppm disappeared while a resonance at 3.4 ppm showed a reduction in signal on SE 135 msec. The ex vivo PMRS confirmed the resonances as mannitol. It was found from the treatment chart that the patient was on i.v. mannitol for intracranial tension reduction until the time of surgery. In three cases of arachnoid cysts only small lactate resonance was observed. In one patient with arachnoid cyst, serum enzyme linked immuno-sorbent assay (ELISA) test for hydatid cyst was positive (Fig. 3), however the diagnosis of arachnoid cyst was made as only lactate resonance was observed on PMRS. In one case the PMRS data were interpreted as inconclusive (Fig. 4) wherein a single resonance at 2.04 ppm which was previously unassigned in cystic lesion was observed. The ex-vivo spectroscopy assigned the resonance as N-acetyl group of compound and the pathology proved to be a glioependymal cyst. In remaining two cases, resonances from choline and lipid/lactate were only observed and were interpreted as representing glioma. However, histopathology proved the cases benign: an infarction and a xanthogranuloma (Fig. 5). In-vivo PMRS correctly identified the pathology in 46/50 (92%) cases in which good quality spectra were obtained. In two cases (4%) MR spectroscopy suggested glioma but

4 106 A. Shukla-Dave et al. / Magnetic Resonance Imaging 19 (2001) Fig. 2. Hydatid cyst. T2W axial image (a) showing a large, well defined, rounded cystic area with hypointense rim and perifocal edema in the right temporo-parietal region. Coronal T1W image (b) shows the cyst to be well defined with mass effect and midline shift. In vivo PMRS from the inset in (b) with STEAM 20 ms(c) shows the resonances of lactate (L) at 1.3ppm, alanine (AL) at 1.5ppm, succinate (S) at 2.4ppm, Choline (C) at 3.22ppm and mannitol (M) at 3.35 and 3.45ppm. SE 135 ms sequence (d) shows phase reversal of lactate and alanine. Note the reduction of the signal seen at 3.35 and 3.45ppm. Ex-vivo PMRS (e and f) confirms the resonances at 3.35 and 3.45ppm are due to mannitol. Patient was on i.v. mannitol infusion until the end of surgery.

5 A. Shukla-Dave et al. / Magnetic Resonance Imaging 19 (2001) histopathology confirmed it to be benign lesions. In the remaining two (4%) cases, the spectral data were inconclusive. Among these two cases one was benign and one malignant lesion. 4. Discussion Fig. 3. Arachnoid Cyst. T1W axial image (a) shows a large cystic lesion in the right frontal region extending to the periphery. Serum ELISA for hydatid cyst was positive in this patient. In vivo PMRS from the inset in (a) with SE 135 ms sequence (b) shows a very small lactate resonance (L) at 1.3ppm. Ex vivo PMRS (c) confirms the presence of lactate resonance. In vivo PMRS is a non-invasive technique to obtain the metabolite profile of normal and abnormal brain. It is widely used as a diagnostic tool in characterization of intracranial mass lesions [3,7,8,11,12,14,15]. Few retrospective studies have shown the utility of this technique in characterization of intracranial cystic lesions [3,8]. The characterization of intracranial cystic lesions in these studies was based on the presence of specific metabolite resonances or a specific combination of known metabolites [3,8,13]. Abscesses characteristically demonstrate resonances from cytosolic amino acids (leucine, isoleucine and valine at 0.9ppm), lactate (1.3ppm), alanine (1.5ppm) and acetate (1.92ppm) with absence of N-acetylaspartate (NAA), Choline (Cho) and Creatine (Cr), which can easily differentiate it from a neoplastic lesion [3,8,10 12]. Recently, it has been shown that the inversion of resonances in the chemical shift range ppm is indicative of an abscess with a specificity rate close to 100% [15]. In the present series, all the 21 cases of abscesses showed resonance at 0.9 ppm. In two cases, abscesses also demonstrated hemorrhage in the wall, simulating hemorrhagic cystic neoplasm. PMRS showed the characteristic resonance at 0.9ppm and the lesions were correctly diagnosed as abscesses, even in the presence of hemorrhage. Neoplasm of the brain characteristically demonstrates increase in choline compounds and lactate with decrease in NAA [12,14,15]. Presence of choline signal in a cystic neoplasm of brain has been reported [3] although a recent series [8] has reported no such finding. In the present series, all except one case of cystic glioma demonstrated choline resonance on in vivo PMRS and were confirmed on ex-vivo PMRS. It is also likely that the voxel may contain the wall of the lesion that may be responsible for the presence of choline in the cystic lesions. In one patient with glioblastoma multiforme, even ex-vivo PMRS did not demonstrate choline signal suggesting that the nonvisualization of the choline resonance was simply not due the sensitivity constraints of the in-vivo PMRS. The non-demonstration of choline in this patient was probably due to resorption of free choline from the wall of cystic glioma. Arachnoid cysts are difficult to be differentiated from a large hydatid cyst and a glioependymal cyst based on conventional MR imaging features alone [12,16]. Arachnoid cysts in the present series demonstrated a very small lac doublet and absence of resonances from NAA, Cr, and Cho at STEAM 20 msec. In one patient with arachnoid cyst, serum ELISA test was positive for hydatid cyst. However, the PMRS did not show any other resonances apart from lactate and was correctly diagnosed as arachnoid cyst in

6 108 A. Shukla-Dave et al. / Magnetic Resonance Imaging 19 (2001) Fig. 4. Glioependymal cyst. T2W axial image (a) shows a large cystic mass in the frontal region in midline with mass effect. Post contrast T1W image (b) shows no obvious enhancement. In vivo PMRS with SE 135 ms shows a resonance (N) at 2.04ppm (c). Exvivo PMRS (d & e) shows a singlet at 2.04ppm assigned as N-acetyl group of compound. Histopathology (f) from the wall of the cyst shows flat cuboidal epithelium resting on a layer of astroglial stroma.

7 A. Shukla-Dave et al. / Magnetic Resonance Imaging 19 (2001) Fig. 5. Xanthogranuloma. T2W axial image through the lateral ventricle (a) shows a hyperintense area with a hypointense rim in the left occipital region with perifocal edema. In vivo PMRS with STEAM 20 (b) shows a very large resonance of lipid at 1.3ppm and 0.9 ppm. AT SE 135 ms (c) sequence there is partial alteration of the phase at 1.3ppm with marked reduction in the signal-suggesting lipid and lactate (Lip/Lac) along with presence of choline (Cho) at 3.22ppm. prospective. Earlier reports of PMRS in arachnoid cysts have also shown similar results [3,8]. Glioependymal cysts are lined by a single layer of ciliated epithelium supported by a delicate connective tissue or glial membrane [16]. These cysts are found both in intraaxial and extraaxial locations have smooth unilocular appearance and are difficult to differentiate from other cystic lesion especially arachnoid cysts when located in the extraaxial location [17]. In the present series, one case of glioependymal cyst was studied. The cyst was extra ventricular, intraaxial mass located in frontal region with a well-defined rim. PMRS showed a single resonance at 2.04 ppm that was previously not observed in a cystic lesion. Since there was no experience of PMRS in glioependymal cysts (of the investigators) in house or in the literature, the study was regarded as inconclusive. Ex-vivo spectroscopy confirmed the resonance as N-acetyl group of compounds. The presence of glial tissue in the ependymal lining may be responsible for the N-acetyl group of compound in this cyst. Three cases of intracranial hydatid cysts in the present series showed lactate, acetate and succinate as reported earlier in literature [13]. One patient with hydatid cyst and pericystic edema showed presence of choline and mannitol on PMRS. The laminated membranes of hydatid cysts are known to be composed of choline containing compounds [18]. Degeneration of the cyst secondary to host response may have resulted in release of free choline and its accumulation in the cyst fluid. In vitro, secondary hydatid cysts have been found to absorb water, electrolytes, mebendazole, cholesterol and certain proteins [19]. It appears that secondary to degeneration, mannitol may have accumulated in the cyst fluid because of increased permeability of the membranes. There is only one report of in vivo MRS of neurocysticercosis in literature [8], which also showed presence of lactate, alanine and succinate similar to hydatid cyst reported earlier [13]. In two benign conditions, infarction and xanthogranuloma, PMRS contributed to a false diagnosis of a malignant lesion. Xanthogranuloma is composed of dense accumulation of xanthoma cells, which are characterized by intracellular accumulation of lipid droplets, predominantly occurring within histiocytes or macrophages [20]. In case of infarct, the area of necrosis is surrounded by an inflammatory response comprising of macrophages and cellular infiltrates. The presence of choline resonance in both these conditions is probably due to inclusion of the wall of these lesions in VOI, which are comprised of dense accumulation of inflammatory cells. Krouwer et al. [21] have reported increase in choline resonance in six patients with nonneoplastic lesions with a dense inflammatory infiltrate. These results suggest that some lesions that prove to be inflammatory, demyelinating, or vascular in nature may demonstrate MR spectroscopic data suggestive of a neoplastic process. Overall, PMRS accurately diagnosed the pathology of cystic lesions in 46/50 cases (92%). In two cases, PMRS did not contribute to the diagnosis and in two cases, falsely

8 110 A. Shukla-Dave et al. / Magnetic Resonance Imaging 19 (2001) diagnosed benign lesions as malignant. In conclusion, PMRS may compliment imaging in preoperatively diagnosing the cystic lesions of the brain and influence the overall management of these lesions. References [1] Kjos BO, Brant-Zawadzki M, Kucharezye W, Kelly Wm, Norman D, Newton TH. Cystic intracranial lesions. Magnetic resonance imaging. Radiology 1985;155: [2] Go KG, Hew JM, Kamman RL, Molenaar Wm, Pruim J, Blaaw EH. Cystic lesion of the brain. A classification based on pathogenesis with consideration of histological and radiological features. Eur J Radiol 1993;17: [3] Poptani H, Gupta RK, Jain VK, Roy R, Pandey R. Cystic Intracranial mass lesions: possible role of in vivo MR spectroscopy in its differential diagnosis. Magn Reson Imaging 1995;13: [4] Desprechins B, Stadnik T, Koertz G, Shaban W, Breuq C, Osteaux M. Use of diffusion weighted MR imaging in differential diagnosis between intracerebral necrotic tumors and cerebral abscesses. AJNR Am J Neuroradiol 1999;20: [5] Aprile I, Iaiza F, Lavaroni A, Budai R, Dolso P, Scott CA, Beltrami CA, Faris G. Analysis of Cystic Intracranial Lesions performed with Fluid Attenuated Inversion Recovery MR imaging. AJNR Am J Neuroradiol 1999;20: [6] Ebisu T, Tanaka C, Umeda M, Kitamura M, Naruse S, Higuchi T, Ueda S, Sato H. Discrimination of brain abscess from necrotic or cystic tumors by diffusion weighted echo planar imaging. Magn Reson Imaging 1996;14: [7] Rand SD, Prost R, Houghton V, Mark L, Strainer J, Johansen J, Kim TA, Chetty VK, Muller W, Meyer G, Krouwer H. Accuracy of single voxel proton MR spectroscopy distinguishing neoplastic from nonneoplastic brain lesions. AJNR Am J Neuroradiol 1997;18: [8] Chang KH, Song IC, Kim H, Han MH, Kim HD, Seong SK, Jung HW, Han MC. In vivo single voxel Proton MR spectroscopy in intracranial cystic masses. AJNR Am J Neuroradiol 1998;19: [9] Demaerel P, Van Hecke P, Van Oostende S, Baert Al, Jacken J, Declercq PE, Eggermont E, Plet C. Bacterial metabolism shown by magnetic resonance spectroscopy. Lancet 1994;344: [10] Harada M, Tanouchi M, Miyushi H, Mishitani H, Kannuki S. Brain abscess observed by localized proton magnetic resonance spectroscopy. Mag Reson Imaging 1994;12: [11] Remy C, Grand S, Lai ES, Belle V, Hoffmann D, Berger F, Esteve F, Ziegler A, LeBas JF, Benabid Al, Decorps M, Segebath CM. 1H MRS of human brain abscesses in vivo, and in vitro. Magn Reson Med 1995;34: [12] Kim SH, Chang KH, Song IC, Han MH, Kim HC, Kang HS, Han MC. Brain abscess and brain tumor: discrimination with in vivo H-1 MR spectroscopy. Radiology 1997;204: [13] Kohli A, Gupta RK, Poptani H, Roy R. In vivo proton magnetic resonance spectroscopy in a case of intracranial hydatid cyst. Neurology 19995;45: [14] Grand S, Passaro G, Ziegler A, Esteve F, Boujet C, Hoffmann D, Rubin C, Segenarth C, Decorps M, Le Bas JF, Remy C. Necrotic Tumour versus Brain Abscess: Importance of Amino Acids detected at 1H MR spectroscopy Initial results. Radiology 1999;213: [15] Poptani H, Gupta RK, Roy R, Pandey R, Jain VK, Chhabra DK. Characterization of intracranial mass lesions with in vivo proton MR spectroscopy. AJNR Am J Neuroradiol 1995;16: [16] Lantos PL, Vandenberg SR Kleihues P. Tumors of the Nervous system. In. Graham DI, Lantos PL Eds. Greenfield s Neuropathology. London Arnold 1997:786. [17] Ismail A, Tampeiri D, Melanson D, Pokrupa R, Villemure JG, Bertrand G. Glioependymal Cysts: CT, and MR Findings. JCAT 1992; 16: [18] Richards KS, Ilderton E, Yardley HJ. Lipids in the laminated layer of liver, lung and daughter hydatid cysts of equine Echinococcus granulosus (Cestoda). Comp Biochem Physiol [B] 1987;86: [19] Schultz KD, Petronio J, Narad C, Hunter SB. Solitary intracerebral juvenile xanthogranuloma. Case report and review of literature. Pediatr Neurosurg 1997;26: [20] Jeffs SA, Arme C. Echinococcus granulosus (Cestoda): uptake of L-amino acids by secondary hydatid cysts. Parasitology 1988;96: [21] Krouwer HGJ, Kim TA, Rand SD, Prost RW, Haughton VM, Ho XC, Jaradeh SS, Meyer GA, Blinadauer KA, Cusick JF, Morris GL, Walsh PR. Single voxel proton MR Spectroscopy of nonneoplastic brain lesions suggestive of a neoplasm. AJNR Am J Neuroradiol 1998;19: