| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Magnetic Resonance Imaging Unit (S.G., F.E., J.F.L.B.), INSERM U438, Université Joseph Fourier, the Corresponding Research Laboratory to the Atomic Energy Center (G.P., A.Z., C.S., C. Rubin, M.D., C. Rémy), the Laboratory of Biochemistry (C.B.), and the Department of Neurosurgery (D.H.), Centre Hospitalier et Universitaire de Grenoble, Hôpital Michallon, 38043 Grenoble cedex 09, France. Received June 23, 1998; revision requested August 5; final revision received March 11, 1999; accepted July 1. Supported in part by the Ligue contre le Cancer, the Association pour la Recherche sur le Cancer, the Ministère de la Santé, and the Biomed 2 program. Address reprint requests to S.G. (e-mail: Jean-Francois.LeBas@ujf-grenoble.fr).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Amino acid concentrations were determined in vitro in 13 purulent samples from brain and nonbrain tissues and in nine aseptic fluids from necrotic brain tumors at two-dimensional (2D) 1H MR spectroscopy and liquid chromatography. Thirty-four patients with cystic intracerebral mass lesions (28 tumors, six abscesses) were examined at 1H MR spectroscopy in vivo.
RESULTS: Amino acids were identified in vitro in both purulent and aseptic samples. Amino acid concentrations measured in the aseptic fluids at both liquid chromatography and 2D MR spectroscopy were far below the detection threshold of in vivo 1H MR spectroscopy. Quantitative results obtained at 2D MR spectroscopy showed no overlap in the ranges of amino acid concentrations in purulent and aseptic samples. In vivo, the proton spectra obtained with a 136-msec echo time (TE) revealed amino acids (inverted peak at 0.9 ppm) in only the abscesses.
CONCLUSION: The detection of amino acid resonance at 0.9 ppm at in vivo 1H MR spectroscopy (136-msec TE) is a promising tool for distinguishing bacterial abscesses and cystic brain tumors.
Index terms: Brain abscess, 10.201, 10.204, 10.256 • Brain neoplasms, 10.36, 10.38 • Magnetic resonance (MR), spectroscopy, 10.12145
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The assignment of various resonances (lactate, acetate, succinate, lipids, alanine, valine, leucine, isoleucine) in the in vivo 1H MR spectra from brain abscesses has been performed by using one-dimensional (1D) and two-dimensional (2D) 1H MR spectroscopy in biopsy samples (4,6,7,12). It has been suggested (7–13) that 1H MR resonances from succinate (2.4 ppm), acetate (1.9 ppm), alanine (1.5 ppm), and from the three amino acids valine, leucine, and isoleucine (0.9 ppm region) are potential abscess markers. Nevertheless, the topic is still controversial (15,16). It has been stated that amino acids could also be detected at in vitro analyses of brain tumor samples, and there have been calls for further in vitro studies (15).
The purpose of this study was to assess whether the resonance lines from amino acids in the in vivo 1D 1H MR spectra provide a specific and sensitive signature for brain abscesses.
We decided, therefore, to proceed along the following lines. (a) To perform in vitro 1D 1H MR spectroscopic measurements in pus samples from various origins, under MR spectroscopic conditions similar to those applied in vivo in brain abscesses (ie, with echo times [TEs] of 136 and 272 msec). The collection of pus samples from various origins allowed the number of observations to be increased substantially. (b) To quantify three amino acids (valine, leucine, and isoleucine) in these samples at 2D 1H MR correlated spectroscopy and biochemical analysis. Quantification by means of 1D 1H MR spectroscopy is difficult because of the overlap in the amino acid multiplet and in lipid resonances. (c) To perform similar 1D and 2D in vitro 1H MR spectroscopic measurements and biochemical analysis with aseptic biopsy samples taken from necrotic cystic brain tumors. (d) To compare the 1D and 2D 1H MR spectra and results from biochemical analysis of the pus samples with those from the tumor biopsy samples, with focus on the amino acid spectral region at 0.9 ppm. (e) To compare the pus spectra obtained at in vitro 1D 1H MR spectroscopy with the brain abscess spectra obtained at in vivo 1H MR spectroscopy.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
In vitro 1H MR spectroscopy.—Each sample was placed in a 5-mm nuclear MR tube without any particular treatment to be as close as possible to in vivo conditions. A capillary tube that contained sodium 3-trimethylsilyl[2,2,3,3-H4]propionate (TSP) at a concentration of 20 mmol/L for the purulent samples or 10 mmol/L for the aseptic samples was used as a reference for external chemical shift and intensity. 1H MR spectroscopy was performed with a spectrometer (AM 400; Bruker, Wissembourg, France) at 37°C. The 1D 1H MR signals were acquired with a spin-echo sequence (4,000/136 and 272 [repetition time msec/TE msec]). The water signal was suppressed by using a low-power presaturation pulse (1,500 msec). The spectral width was 4,000 Hz, and the data comprised 16,000 points.
The data were Fourier transformed without any filtering. The 2D spectra were obtained with a standard 2D correlated spectroscopic sequence with water suppression. The spectral width in both dimensions was 4,000 Hz. The number of complex data points was 512 in the t2 dimension and 256 in the t1 dimension. The number of averages obtained was 16 for each t1 step. Data were filtered with sine-bell windows, were zero-filled in both dimensions, and were then Fourier transformed.
The 2D correlated spectroscopic maps were presented in magnitude mode. The resonance peaks of the 1D spectra were assigned on the basis of previous results (17). For semiquantitative analysis, correlation peaks for the main amino acid resonances (valine [
1;2 = 1.02;2.28], leucine [1;2 = 0.96;1.72], isoleucine [1;2 = 0.96;1.99]) were integrated and normalized with respect to the TSP peak volume. A further correction was applied to take into account the difference in TSP concentrations in the purulent and aseptic samples.
Ion exchange amino acid chromatography.—Amino acids were analyzed at liquid chromatography in 18 (11 purulent and seven aseptic samples) of the 22 samples. The samples were centrifuged, and supernatants were deproteinated with sulfosalicylic acid powder (to a final concentration of 40 mg/mL). The deproteinated supernatants were diluted with the appropriate Beckman special buffer, Li S (pH 2.2), to the ratio of 1:5 (1:20 for the most concentrated ones) and were analyzed. The apparatus (Beckman Coulter Instruments, Palo Alto, Calif) included a 10-cm–long cation-exchange resin column and a ninhydrin-derivative colorimetric detector adjusted to 570 or 440 nm; these were used with the four eluting buffers for biologic fluids—Li A, Li D, Li E, and Li F. Quantification was performed according to the Beckman-System-Gold protocol.
In Vivo Studies
Patients.—Thirty-four patients (13 women and 21 men; mean age, 51.5 years; age range, 18–74 years) who presented with neurologic symptoms and who had an intracranial necrotic mass lesion on MR images were examined at in vivo 1H MR spectroscopy prior to biopsy and/or surgery. The MR images were interpreted by a neuroradiologist (J.F.L.B. or S.G.). All lesions exhibited a necrotic center that was hypointense on T1-weighted images and hyperintense on T2-weighted images. The perilesional ring was regular or moderately irregular and enhanced after an injection of gadolinium-based contrast material (Dotarem [gadoterate meglumine]; Guerbet, Roissy, France; 0.1 mmol per kilogram of body weight, intravenous injection), which was administered after 1H MR spectroscopy. The diameter of the necrotic cystic center was at least 2 cm.
All patients required hospitalization, but only one previously had clinically important medical problems. This patient had been treated with radioactive implants for a low-grade astrocytoma, which subsequently developed into a necrotic cystic lesion. The patients who had malignant metastases were not known to have had cancer previously.
Results of serologic testing for the human immunodeficiency virus were negative in all patients with abscesses. One patient with a cerebral abscess was receiving antibiotics for facial cellulitis. (The other patients with abscesses were not taking antibiotics.) Twenty-one lesions were high-grade gliomas (grades III–IV), seven were metastases, and six were abscesses. The diagnoses were based on the histopathologic findings in stereotactic biopsy samples or in surgical specimens. In the cases with abscess, the etiologic agents were identified by means of culture study of the pus sample. The culture of the pus samples showed a variety of causal organisms, as given in Table 1. The study was performed with the approval of the local ethics committee, and informed consent was obtained for each patient.
In vivo 1H MR spectroscopy.—1H MR spectroscopic studies were performed as previously described (7). Before MR spectroscopy, transverse T2-weighted MR scout images were obtained (3,900/100). Two water-suppressed, double spin-echo 1H MR spectroscopic imaging sequences were performed (1,500/136 and 272). The acquisition time was 8 minutes for each sequence. The digital resolution of the spectroscopic imaging sequence was 1 cm in the anteroposterior or left-right directions and 2 cm in the caudocranial direction. Thus, for voxels centered in the necrotic cystic part of the lesion (2 cm in diameter), the spectra was slightly contaminated by surrounding tissues. After image acquisition at spectroscopy, T1-weighted images were acquired (550/15) in two orthogonal planes with and without gadolinium enhancement.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
With this TE, the resonance lines from amino acids and lipids centered at 0.9 ppm exhibited opposite signs. This was also the case with the resonance lines from lactate and lipids centered at 1.3 ppm. Lipids were identified in 11 of the 13 purulent samples and in four of the nine samples from aseptic necrosis. All samples (purulent and aseptic) showed the 1.3-ppm resonance of lactate. The multiplet at 0.9 ppm due to the amino acids was clearly visible in all purulent samples from abscesses and was hardly detectable in aseptic samples from brain tumors.
Figure 2 shows the 2D correlated spectroscopic maps obtained in pus samples from a brain abscess and an abdominal abscess and in an aseptic biopsy sample from a necrotic cystic tumor. These spectra show that valine, leucine, and isoleucine are the three amino acids that gave rise to the inverted 0.9-ppm resonance on the 1D spectra of the purulent samples. The cross peaks of these three amino acids were well resolved on the 2D correlated spectroscopic maps (Fig 2), which allowed their quantification (Fig 3, Table 2). Quantitative analysis indicated that the intensity of the amino acid cross peaks was much higher in purulent fluids than in the aseptic fluids (Fig 3, Table 2). These results were confirmed at liquid chromatographic analysis (Fig 4, Table 3).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Brain abscess.—The resonances of amino acids (0.9 ppm), acetate (1.9 ppm), and succinate (2.4 ppm) were identified, respectively, in six, six, and four of the six abscesses analyzed (Fig 7). Lactate was always detected (1.3 ppm), and lipids (0.9 and 1.3 ppm) were found in three patients.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
To our knowledge, results from all in vivo 1H MR spectroscopic studies of brain abscesses reported so far have exhibited a group of resonance lines centered at 0.9 ppm (Table 4). The 20-msec and 270-msec TEs do not allow separation of the resonances from amino acids (––CH3 of valine, leucine, and isoleucine) and lipids (––CH3) within this group. This separation requires the acquisition of spectra with a TE of 136 msec (4,7,10). The precise identification of the 0.9-ppm resonance is particularly important, as the spectra from metastatic brain tumors and high-grade glioblastomas often exhibit resonances centered at this frequency (18–21). On the basis of its behavior with a 136-msec TE, the 0.9-ppm resonance is, in those cases, generally assigned to lipids.
Table 4 shows that the inversion of the 0.9-ppm resonance—an indication of the presence of the amino acids—was found in the spectra for 19 of 20 (95%) abscesses. The sensitivity might have been close to 100% (20 of 20); as in the exceptional case, the spectral quality of the 136-msec–TE spectrum appeared insufficient to differentiate the 0.9-ppm peaks from lipids and amino acids (11). Recently, the specificity of the amino acid resonances for the diagnosis of brain abscesses has been questioned (15). It was stated that amino acids are not specific features of abscess fluids since these compounds can be detected in vitro in brain tumors (15,22).
In this study, we have, therefore, focused on the in vivo MR spectroscopic visibility and on the concentrations of the three amino acids valine, leucine, and isoleucine in the samples taken from brain abscesses and cystic tumors. Concentrations were determined at in vitro MR spectroscopy and at biochemical analysis. As brain abscesses are rare, purulent samples were also taken from outside the central nervous system to increase the number of observations.
The concentrations of amino acids were expected to be similar in both types of purulent samples, as they were due to the proteolytic activity of the polymorphonucleocytes (23,24). The in vitro MR spectroscopic measurements were carried out in samples that had undergone no particular treatment (except freezing and thawing). Thus, the conditions for in vitro MR spectroscopic measurement were very similar to those encountered in vivo. Our results indicate that the amino acids are detectable in vitro in brain tumor samples, which is in agreement with the results of Silberstein (15) and Silberstein and Dodd (22). The in vivo MR spectroscopic visibility of amino acids in the 0.9-ppm region with a 136-msec TE is, however, another issue.
The mean concentrations of valine and leucine found in this study in the purulent samples by means of liquid chromatography (8.16 mmol/L) was higher than the values reported by Schumacher et al (25) in 14 abscess fluid samples taken from outside the central nervous system (2.57 mmol/L) and by Martinez-Pérez et al (12) in two brain abscesses (4.07 mmol/L). These differences may be due to differences in the techniques for measurement, in the experimental conditions, and in the sources of samples. For instance, Schumacher et al (25) determined the mean concentrations of valine and leucine by comparing the integrals of the resonances in the 0.95–1.06-ppm region of a 134-msec–TE spectrum with those obtained after the addition of known amounts of valine and leucine. This approach may have resulted in an underestimation of the amino acid concentration because of an overlap with noninverted components such as lipids (7,12).
In seven of the nine aseptic fluid samples, biochemical analysis revealed concentrations of valine, leucine, and isoleucine that were well below 1 mmol/L. These concentrations are very likely below the concentration threshold required for in vivo MR spectroscopic visibility with a 136-msec TE. On average, the concentrations of valine, leucine, and isoleucine were, respectively, 20, 50, and 78 times higher in the purulent samples than in the fluids from cystic brain tumors.
In addition to the concentrations, the relaxation times also have to be taken into consideration when in vivo MR spectroscopic signal intensities are compared. As a 2D correlated spectroscopic experiment is basically a spin-echo experiment with variable TE, one may consider that cross peak volumes are better predictors of the amplitudes of resonance lines obtained with in vivo spin-echo MR spectroscopic experiments than are the concentrations obtained at biochemical analysis. Cross peaks from valine and leucine were at the limit of detectability in the 2D spectra from aseptic fluids, and the cross peak from isoleucine was undetectable. For valine, the ratio of the mean normalized cross peak in the spectra from purulent and aseptic fluids was approximately 50. It is noteworthy that there was no overlap between the ranges in amino acid concentrations in the purulent and aseptic samples.
These results strongly suggest that the signal-to-noise ratio on in vivo 1H MR spectra from necrotic brain tumors is insufficient for the detection of valine, leucine, and isoleucine. This also explains why the inversion of resonances in the 0.9-ppm region was never reported, to our knowledge, in brain tumors despite the large number of 1H MR spectroscopic studies that were performed with a 136-msec TE. The inversion of resonances in the chemical shift range (0.9–1.1 ppm) is, thus, indicative of an abscess with a specificity rate close to 100%.
The sensitivity rate depends on the in vivo visibility of the inverted resonances in the 0.9–1.1-ppm range with a 136-msec TE. As mentioned earlier, this sensitivity rate nears 100% when it is evaluated on all 136-msec 1H MR spectra from intracerebral abscesses that have been reported so far (Table 4). However, the specificity and sensitivity rates have to be confirmed with data from a larger number of patients.
It is important to emphasize that the detection of amino acids in brain abscesses applies only to bacterial infection. In cases of parasitic abscesses (one hydatid cyst reported by Kohli et al [26], 11 toxoplasmic abscesses and four cryptococcomas reported by Chang et al [27]), no amino acids were detected. Similarly, in a fungal psoas abscess, amino acids were at low concentrations (25). Consequently, 1H MR spectroscopy is likely to be unhelpful in the diagnosis of parasitic abscesses. This is unfortunate, as it is thus impossible, at present, to make a differential diagnosis between toxoplasmosis and lymphoma by means of 1H MR spectroscopy in patients with acquired immunodeficiency syndrome.
In summary, findings from this study have demonstrated that the detection of amino acid resonances at 0.9 ppm at in vivo 1H MR spectroscopy permits the differentiation between cystic brain tumors and intracranial bacterial abscesses.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, C. Rémy, J.F.L.B.; study concepts, S.G., C.S., J.F.L.B., M.D.; study design, M.D., A.Z., S.G.; definition of intellectual content, J.F.L.B., M.D., C.S., S.G.; literature research, C. Rémy, S.G.; clinical studies, S.G., F.E., D.H.; experimental studies, G.P., S.G., A.Z., C.B.; data acquisition, G.P., S.G., A.Z., C.B.; data analysis, G.P., C. Rémy, M.D., S.G., A.Z.; statistical analysis, C. Rubin, G.P., S.G., F.E.; manuscript preparation and review, S.G., C.R., M.D., C.S.; manuscript editing, S.G., C. Rémy.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. N. Al-Okaili, J. Krejza, S. Wang, J. H. Woo, and E. R. Melhem Advanced MR Imaging Techniques in the Diagnosis of Intraaxial Brain Tumors in Adults RadioGraphics, October 1, 2006; 26(suppl_1): S173 - S189. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Himmelreich, R. Accurso, R. Malik, B. Dolenko, R. L. Somorjai, R. K. Gupta, L. Gomes, C. E. Mountford, and T. C. Sorrell Identification of Staphylococcus aureus Brain Abscesses: Rat and Human Studies with 1H MR Spectroscopy Radiology, July 1, 2005; 236(1): 261 - 270. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. H. Lai, K. T. Li, S. S. Hsu, C. C. Hsiao, C. W. Yip, S. Ding, L. R. Yeh, and H. B. Pan Pyogenic Brain Abscess: Findings from In Vivo 1.5-T and 11.7-T In Vitro Proton MR Spectroscopy AJNR Am. J. Neuroradiol., February 1, 2005; 26(2): 279 - 288. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Garg, R. K. Gupta, M. Husain, S. Chawla, J. Chawla, R. Kumar, S. B. Rao, M. K. Misra, and K. N. Prasad Brain Abscesses: Etiologic Categorization with in Vivo Proton MR Spectroscopy Radiology, February 1, 2004; 230(2): 519 - 527. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. K. Lehtimaki, P. K. Valonen, J. L. Griffin, T. H. Vaisanen, O. H. J. Grohn, M. I. Kettunen, J. Vepsalainen, S. Yla-Herttuala, J. Nicholson, and R. A. Kauppinen Metabolite Changes in BT4C Rat Gliomas Undergoing Ganciclovir-Thymidine Kinase Gene Therapy-induced Programmed Cell Death as Studied by 1H NMR Spectroscopy in Vivo, ex Vivo, and in Vitro J. Biol. Chem., November 14, 2003; 278(46): 45915 - 45923. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Bulakbasi, M. Kocaoglu, F. Ors, C. Tayfun, and T. Ucoz Combination of Single-Voxel Proton MR Spectroscopy and Apparent Diffusion Coefficient Calculation in the Evaluation of Common Brain Tumors AJNR Am. J. Neuroradiol., February 1, 2003; 24(2): 225 - 233. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Majos, J. Alonso, C. Aguilera, M. Serrallonga, J. J. Acebes, C. Arus, and J. Gili Adult Primitive Neuroectodermal Tumor: Proton MR Spectroscopic Findings with Possible Application for Differential Diagnosis Radiology, November 1, 2002; 225(2): 556 - 566. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. H. Lai, J. T. Ho, W. L. Chen, S. S. Hsu, J. S. Wang, H. B. Pan, and C. F. Yang Brain Abscess and Necrotic Brain Tumor: Discrimination with Proton MR Spectroscopy and Diffusion-Weighted Imaging AJNR Am. J. Neuroradiol., September 1, 2002; 23(8): 1369 - 1377. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. K. Gupta, D. K. Vatsal, N. Husain, S. Chawla, K. N. Prasad, R. Roy, R. Kumar, D. Jha, and M. Husain Differentiation of Tuberculous from Pyogenic Brain Abscesses with In Vivo Proton MR Spectroscopy and Magnetization Transfer MR Imaging AJNR Am. J. Neuroradiol., September 1, 2001; 22(8): 1503 - 1509. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Himmelreich, T. E. Dzendrowskyj, C. Allen, S. Dowd, R. Malik, B. P. Shehan, P. Russell, C. E. Mountford, and T. C. Sorrell Cryptococcomas Distinguished from Gliomas with MR Spectroscopy: An Experimental Rat and Cell Culture Study Radiology, July 1, 2001; 220(1): 122 - 128. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||
], other [
]) and aseptic (
) samples. Levels are expressed in arbitrary units and were determined by means of integration of the correlation peaks obtained at 2D correlated spectroscopy and by means of normalization to TSP. There is no overlap between tumors and abscesses.
