| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neuroradiology |
1 From the Depts of Radiodiagnosis (R.K.G.), Radiology (M.G., R.K.G., S.C., R.K.), and Microbiology (K.N.P.), Sanjay Gandhi Post-Graduate Institute of Med Sciences, Lucknow-226014, India; Dept of Neurosurgery, King George’s Med College, Lucknow, India (M.H., J.C.); Dept of Mathematics, Indian Institute of Technology, Kanpur, India (S.B.R.); and Dept of Biochemistry, Univ of Lucknow, India (M.K.M.). Received Oct 15, 2002; revision requested Dec 23; final revision received May 15, 2003; accepted Jun 16. S.C. was supported by financial assistance from the Council of Scientific and Industrial Research, New Delhi, India. Address correspondence to R.K.G. (e-mail: rakeshree@hotmail.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: MR imaging and in vivo single-voxel proton MR spectroscopic data obtained from 75 patients with brain abscesses were retrospectively analyzed. Ex vivo spectroscopic experiments with the pus from 45 of these patients also were performed, and the data were further categorized on the basis of bacteriologic information. Quantification of various metabolites and metabolite ratios and statistical analyses of lactate and lactate/amino acid (AA) ratio levels were performed by using one-way analysis of variance.
RESULTS: On the basis of in vivo proton MR spectroscopic and bacteriologic analysis findings, data were classified into three categories: Group 1 data showed resonances of lactate, AAs, and acetate, with or without succinate at proton MR spectroscopy; cultures for this group showed obligate anaerobes or a mixture of obligate and facultative anaerobes. The metabolite patterns in the group 2 and group 3 data were similar to the pattern of the group 1 data, with the exception that acetate and succinate resonances were absent. Culture was positive for either obligate aerobes or facultative anaerobes in group 2 and was sterile in group 3. At analysis of variance, in vivo data showed significant differences in lactate/AA ratios (P = .008), and ex vivo data showed significant differences in lactate levels (P = .001) among the three groups.
CONCLUSION: It is possible to differentiate anaerobic from aerobic or sterile brain abscesses on the basis of metabolite patterns observed at in vivo proton MR spectroscopy. This information may be useful in facilitating prompt and appropriate treatment of patients with these abscesses.
© RSNA, 2003
Index terms: Brain, abscess, 13.256 • Brain, infection, 13.201, 13.202, 13.204, 13.256 • Brain, MR, 13.121411, 13.121412, 13.121415, 13.12145 • Magnetic resonance (MR), spectroscopy, 13.12145
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Computed tomography and conventional magnetic resonance (MR) imaging are the main noninvasive modalities used to diagnose brain abscess. However, these modalities have limitations because the imaging features of brain abscess are nonspecific and may simulate those of cystic rim-enhancing mass lesions of varying etiologies. Proton MR spectroscopy complements conventional MR imaging by enabling better lesion characterization (5,6) and is increasingly being used in neurosurgical practices to help characterize cystic intracranial lesions that have similar appearances on MR images (7). Proton MR spectroscopy is being used to differentiate brain abscess from glioblastoma multiforme (8–10). More recently, this examination has been used successfully to discriminate tuberculous from pyogenic abscesses (11).
While commenting on the study of Dev et al (7), C. S. Zee hypothesized that brain abscesses could be differentiated into anaerobic and aerobic types by using proton MR spectroscopy; however, to our knowledge, before the present study, there had been no published data on such differentiation in the literature.
The purpose of our study was to compare the metabolite patterns observed at in vivo proton MR spectroscopy of brain abscesses in patients for whom bacteriologic information was obtained from cultures and to categorize the MR spectral patterns with respect to the underlying etiologic agents.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The source of the primary infection could be identified in the majority of the patients. The primary infection was otitis media in 21, sinusitis in 14, dental caries in seven, septicemia in eight, due to postoperative complications in three, and due to posttraumatic complications in five patients; fever was the most common clinical symptom. However, in 17 (23%) of the 75 patients, no apparent primary focus of infection could be identified, even after thorough clinical and laboratory examinations. Twenty-two patients were receiving prolonged antibiotic therapy at the time of the study.
Ultrasonographically guided aspiration of the abscesses was performed in all cases, and the pus collected was divided into two portions. One portion of the pus was snap frozen in liquid nitrogen to prevent any metabolite changes that might influence the results of ex vivo spectroscopy. The rest of the pus was sent to be cultured immediately after its aspiration for detailed bacteriologic study. Ex vivo spectroscopic experiments were performed with the pus collected from 45 (60%) of the 75 patients. We could not collect pus samples from the remaining 30 patients because they underwent emergency surgery owing to their deteriorating clinical condition and no provisions for snap freezing samples were made at the collaborating institution during these emergencies. Therefore, ex vivo spectroscopy could be performed for only 45 patients.
Conventional MR Imaging, MR Spectroscopy, and Analysis
All 75 patients with brain abscesses underwent conventional MR imaging and in vivo proton MR spectroscopy in a single session. MR imaging and MR spectroscopy were performed with a 1.5-T superconducting system (Magnetom SP; Siemens, Erlangen, Germany) by using a circularly polarized head coil. MR imaging was performed with routine proton density–weighted, T2-weighted (2,200/20, 80 [repetition time msec/echo time msec], one signal acquired), and T1-weighted (660/15, two signals acquired) spin-echo sequences. T1-weighted MR imaging was also performed after injecting 0.1 mmol of gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) per kilogram of body weight at the end of in vivo spectroscopy. An experienced radiologist (R.K.G.) reviewed the MR imaging data on all 75 patients to determine the topographic location of their brain abscesses and identify the characteristic appearance of the abscesses: a hyperintense core with a hypointense rim on T2-weighted images, a hypointense core with an isointense to slightly hyperintense rim on T1-weighted images, and rim enhancement on postcontrast T1-weighted images.
Single-voxel volume-selective water-suppressed proton MR spectroscopy was performed by using stimulated-echo acquisition mode (STEAM) and spin-echo sequences. We initially performed both STEAM and spin-echo sequences (in 43 patients); however, only the STEAM sequence was performed in 11 patients because these patients were too sick to endure the extended period required for both sequences. We believed that the spin-echo sequence was sufficient for abscess characterization because lipids with short T2 values are suppressed at this examination. Thus, we performed only the spin-echo sequence in the remaining 21 patients. Overall, 64 (85%) of the 75 patients underwent spin-echo MR spectroscopy.
A voxel of 4.09–8.00 cm3, depending on the size of the abscess, was placed in the abscess cavity, with the aim of avoiding spectral contamination with either the surrounding brain parenchyma or the wall of the abscess cavity. To maintain uniformity, one author (R.K.G.) placed the voxel for spectroscopy in all patients. When multiple abscesses were present, the largest abscess was used for voxel placement for spectroscopy. The experimental parameters used to perform STEAM MR spectroscopy were 3,000/20, 30-msec mixing time, and 128 acquisitions. The experimental parameters used to perform spin-echo MR spectroscopy were 3,000/135 and 128 acquisitions. The spin-echo sequence was used to identify the resonances with short T2 values and to observe and confirm the phase reversal shown by the J-coupled multiplets (5). After global shimming, voxel shimming was performed, and a full width at half maximum of 4–6 Hz was achieved in the majority of cases.
Two spectroscopists (M.G., R.K.G.) independently analyzed and quantified the in vivo spin-echo MR spectroscopic data by using a standard JAVA-based version of MR imaging user interface signal-processing software (JAVA-based Magnetic Resonance User Interface-Europe Project). The spectral data would be categorized on the basis of findings in the bacterial cultures. To quantify the metabolites, we selected the good-quality (full width at half maximum, 4–6 Hz) spin-echo MR spectra of 58 (91%) of 64 patients. We chose the spin-echo rather than STEAM sequence for quantification because with spin-echo sequences, metabolites with short T2 values, including lipids that overlap with lactate (Lac) and amino acid (AA) peaks at 1.33 and 0.90 ppm, respectively, get suppressed, and this suppression results in the proper quantification of the remaining metabolites. Although the data from the remaining six patients were of sufficient quality for identification of the metabolite peaks, poor full widths at half maximum (ie, >8 Hz) forced us to exclude these spectra owing to the possibility of erroneous quantification. After a Hankel Lanczos singular values decomposition filter was applied, free induction decay was zero filled to 4,000 data points, and an exponential multiplication was applied before Fourier transformation (12).
Baseline correction of all spectra was performed before quantification. The typical advanced method used to create an accurate, robust, and efficient spectral-fitting algorithm based on prior knowledge of the metabolites visible on the spectrum was applied for quantification. The signal resonance of Lac at 1.33 ppm was used as the internal chemical shift reference. After quantifying the peak areas of various metabolites, we calculated the following peak area ratios: Lac/AA, Lac/acetate (Ac), Lac/succinate (Suc), and Ac/Suc.
Culture of Pus Samples
The pus immediately after being aspirated was cultured in media (BACTEC NR 730 media 16A and 17A; Becton Dickinson, Sparks, Md) to isolate the aerobic and anaerobic bacteria. The inoculated media were incubated at 37°C, and growths were monitored by using the BACTEC NR 730 system every day for 7 days. The media with positive growth values were subcultured on appropriate solid media and incubated both aerobically and anaerobically at 37°C. Anaerobic incubation in jars filled with a gas mixture (N2 80%–90%, CO2 5%–10%, H2 5%–10%) was performed by using the Anoxomat system (Mart Microbiology BV, Lichtenvoorde, the Netherlands). All bacterial isolates were identified after standard biochemical tests were performed. In addition to culture sampling, Gram stains of all pus samples were routinely performed.
Ex Vivo Proton MR Spectroscopy of Pus
Ex vivo proton MR spectroscopy was performed in 45 (60%) of the 75 cases to reconfirm the findings seen at in vivo MR spectroscopy and to compare the in vivo quantification results with the ex vivo data. Ex vivo spectroscopy was performed during 4 hours of pus collection. The pus samples were prepared after mixing 450 µL of pus with 50 µL of deuterium oxide to yield a total volume of 500 µL. Peaks were referenced to sodium salt of 3-(trimethylsilyl) 2,2,3,3-d4-propionate (TSP) in deuterium oxide (0.75% wt/vol) as an external standard. The ex vivo experiments were performed at 25°C by using a spectrometer (model 300 DRX; Bruker, Fallanden, Switzerland) equipped with a vertical 7.1-T, 54-mm-bore magnet operating at 300.13-MHz frequency for protons by using a 5-mm multinuclear inverse-detection probe head with a shielded z gradient. The spectroscopic task was controlled by using a Silicon Graphics Indy workstation (Bruker, Fallanden, Switzerland), and spectral plotting was performed by using computer software (XWINNMR, version 2.6; Bruker).
Standard one-dimensional experiments were performed by using two sequences: a single-pulse–collect spectrum sequence with frequency-selective water presaturation and a Hahn spin-echo sequence with an echo time of 160 msec to observe the phase reversal of the J-coupled multiplets. The typical parameters used to perform single-pulse spectroscopy were 32,000 data points, a spectral width of 3,500 Hz, 128 acquisitions, a flip angle of 90°, and a relaxation delay of 5 seconds.
Standard two-dimensional correlation spectroscopy was also performed with frequency-selective water presaturation. The spectral width was 3,591 Hz in both directions, with 256 evolution time increments and a 2-second recycle delay. For each evolution time step, 16 transients were collected with 1,000 data points. The data were weighted with a sine-bell window function in both dimensions before double-Fourier transformation. The final data matrix was 1,024 x 256 points. The ex vivo assignment of signal resonances was performed with respect to the chemical shift position of TSP centered at 0.0 ppm by using values reported in the literature (13,14).
Absolute Quantification of Metabolites at ex Vivo Spectroscopy
Quantification of the various metabolites visible on the ex vivo MR spectrum was performed to confirm the in vivo spectroscopic findings. Similar to the in vivo data, the ex vivo data were analyzed and quantified by two authors (M.G., S.C.) independently. Phase and baseline corrections were performed before the integration of TSP and all the other metabolites. One-dimensional single-pulse spectra were considered for the integration. At 0.90 and 1.33 ppm, AAs and Lac, respectively, overlap with the terminal methyl (
CH3) and methylene (CH2) groups of lipids at the same chemical shift positions that can result in the improper estimation of these metabolites. Thus, the signal resonances of valine and isoleucine at 3.62 and 3.67 ppm, respectively, and the quartet for Lac at 4.11 ppm—instead of the signal resonances of AAs and Lac at 0.90 and 1.33 ppm, respectively—were used for the integration. The integration of leucine on the one-dimensional spectrum was not performed because this metabolite was not clearly visible; however, it was evident in the two-dimensional correlation spectroscopy experiment. Quantification was performed by using standard software (NMRQUANT, version 2.6 [part of 300 DRX spectrometer]; Bruker) available with the spectrometer. The concentrations of various metabolites, in millimoles per liter, were computed with respect to the known concentration of TSP in the capillary tube.
The concentrations of various metabolite ratios–specifically, Lac/AA, Lac/Ac, Lac/Suc, and Ac/Suc–as were used for in vivo quantification, also were calculated. On the in vivo spin-echo spectrum, the signal resonance at 0.90 ppm was used as the AA peak; however, at ex vivo spectroscopy, for AAs, combined concentrations of valine at 3.62 ppm and isoleucine at 3.67 ppm were obtained by adding the individual concentrations of these metabolites.
Statistical Analyses of in Vivo and ex Vivo Spectroscopic Data
To compute the minimum quantity of in vivo spectroscopic data required to perform statistical analyses, power analysis was performed with standard software (Power and Sample Size [PASS] Analysis, version 6.0; NCSS, Kaysville, Utah). The Lac/Ac ratio was used as a key variable for establishing the difference among groups. We assumed that in anaerobic abscesses the mean Lac/Ac ratio would be 0.5 ± 0.4 (SD). In aerobic abscesses, the ratio was close to zero, but we assumed this value to be 0.1 ± 0.4. For an 
(probability of type I error) of .05 and a ß (probability of type II error) of .2, the minimum required sample size was calculated to be 16 for each category at two-tailed analysis.
Because Ac, as compared with Suc, was present in all of the anaerobic samples, we used mean ratios of Lac with Ac. To compare the mean Lac/AA ratios among the different categories of in vivo spectroscopic data, one-way analysis of variance was performed. For the sample size of 16, the power was found to be 91.8%. The difference in mean Lac/AA ratio and Lac levels in the ex vivo study was also tested for multiple comparisons by using statistical software (SPSS, version 10.0 [standard version]; SPSS, Chicago, Ill). This was done to look for any significant difference in Lac and Lac/AA ratio among different categories of abscesses based on their etiologies. P values of less than .05 were considered to be statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
We found that Suc was always present in conjunction with Ac but never present alone in any of the patients in our study. Six (26%) of the 23 patients in group 1 had Ac resonances in conjunction with resonances of the above-mentioned metabolites, except Suc, at in vivo spectroscopy; however, the bacteriologic findings of the abscesses in these patients were unaffected by the presence or absence of Suc. The B fragilis group was seen in the cultures from the pus, irrespective of the presence or absence of Suc. The in vivo spectra for the 30 (40%) patients in group 2 and the 22 (29%) patients in group 3 showed metabolite patterns that were similar to the pattern for the group 1 patients, with the exception that signal resonances of Ac and Suc were consistently absent in groups 2 (Fig 2c) and 3.
The results of ex vivo spectroscopy of the pus from 45 patients confirmed the findings seen in vivo. In 17 (38%) patients from group 1, in addition to the in vivo signal resonances of the metabolites just described, additional peaks of the glutamate-glutamine complex at 2.09–2.36 ppm, of lysine at 1.73 and 3.01 ppm, of taurine at 3.24 and 3.42 ppm, of valine at 3.62 ppm, and of isoleucine at 3.67 ppm were observed on the one-dimensional single-pulse spectrum (Fig 1c). Lipid peaks at 0.90, 1.30, 1.58, 2.02, 2.24, and 2.80 ppm assigned to terminal methyl (CH3), acyl chain (CH2)n, OCCH2CH2, CH2
CH, COCH2, and CHCH2CH of the fatty acyl chain, respectively, were clearly observed on the spectrum.
On the spin-echo MR spectra, peaks of AAs (ie, valine, isoleucine, and leucine), Lac, and alanine showed phase reversal, whereas the triplets of taurine and the signal resonances of the rest of the metabolites remained unchanged. At two-dimensional correlation spectroscopy analysis, separate contours of valine, isoleucine, Lac, alanine, lysine, the glutamate-glutamine complex, taurine, and lipids showed correlations with their respective counterparts (Fig 1d).
For 13 (29%) patients from group 2 (Fig 2d) and 15 (33%) patients from group 3, the ex vivo spectra showed metabolite patterns similar to that in group 1, with the exception that resonances of Ac and Suc were absent.
Quantification and Statistical Considerations
The results of quantitative data analyses of the various metabolites and metabolite ratios are given in Tables 2–4. The peak area ratios of various metabolites, expressed as means ± SDs, and the statistical results for 58 of the 64 patients in whom quantifiable in vivo spin-echo MR spectra were obtained are given in Table 2. The absolute concentrations of various metabolites and their ratios in 45 patients, also expressed as means ± SDs, along with the statistical results of the ex vivo MR spectroscopic examinations, are summarized in Tables 3 and 4. As evident from the data in Table 2, the mean Lac/AA peak area ratio was lowest in the group 1 patients (ie, with anaerobic abscesses), slightly higher in the group 3 patients (ie, with sterile abscesses), and highest in the group 2 patients (ie, with predominantly aerobic abscesses).
|
|
|
The Ac/Suc ratio calculated in vivo (Table 2) revealed that there were higher concentrations of Ac than Suc; these findings were confirmed at ex vivo spectroscopy of individual concentrations of these metabolites and their ratios (Tables 3 and 4). However, the in vivo and ex vivo spectroscopic data did not reveal consistent results regarding Lac/Ac ratios.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Gas liquid chromatography has been used to rapidly differentiate anaerobes from aerobes in clinical samples on the basis of the presence of volatile and nonvolatile fatty acids on the chromatogram (18,19). In the present study, we attempted to categorize brain abscesses on the basis of their etiologies by using the metabolite information available from noninvasive in vivo spectroscopic examination.
The literature contains a number of studies in which in vivo, ex vivo, and in vitro proton MR spectroscopy was performed with or without corresponding bacteriologic information (7–12,20–26). However, to our knowledge, before this investigation, there had been no study in which the metabolite information was directly correlated with the type of bacteria inhabiting the abscesses. The lack of published correlations between the metabolite pattern and the etiology has probably been due to the small sample sizes in the majority of studies in the literature (12,20–23).
In the current study, the selective presence of Suc and/or Ac signal resonances at in vivo MR spectroscopy of abscesses in the group 1 patients suggests that these resonances act as signatures for anaerobic infection. These metabolite resonances are probably the result of enhanced glycolysis and the fermentative pathways for energy generation (8). Results of studies focused on bacterial metabolism suggest that glycolysis is the universal cycle, irrespective of the aerobic or anaerobic nature of the bacteria. The pyruvate from glycolysis may enter into the tricarboxylic acid cycle in aerobes and in facultative anaerobes in the presence of oxygen. Alternatively, the pyruvate may follow the anaerobic fermentative route in anaerobes, where it may undergo carboxylation to form Suc via oxaloacetate, malate, and fumarate; and part of the pyruvate also forms Ac.
The involvement of alternative anaerobic pathways for energy demands during Ac and Suc production by anaerobic bacteria may cause the appearance of these resonances in anaerobic abscesses (group 1). On the contrary, the absence of Ac and Suc in the spectra of the group 2 patients (with aerobes and facultative anaerobes) can be explained by the fact that the pyruvate from glycolysis follows the tricarboxylic acid cycle. Although Suc is one of the intermediate metabolites of the tricarboxylic acid cycle, Suc does not accumulate during its formation and immediately transforms to the next intermediate compound; this phenomenon results in the absence of Suc in the spectra of group 2 (27). Suc was earlier described as a marker for anaerobic infections in clinical samples at gas liquid chromatography (18).
The spectroscopic findings in the present study appear to be consistent with earlier findings. Demaerel et al (20) in their study of a single case observed both Ac and Suc corresponding to cultures showing Propionibacterium acnes and Haemophilus aphrophilus (20). Rémy et al (24) also reported signal resonances of Suc and/or Ac in four cases of mixed obligate anaerobe and facultative anaerobe infections. These findings are consistent with our cases of mixed anaerobe and facultative anaerobe infections. Kim et al (8) and Martínez-Pérez et al (12) described the spectroscopic and culture data in seven and two cases of brain abscesses, respectively. In two cases in the Kim et al study and in one case in the Martínez-Pérez et al study, signal resonances of both Ac and Suc were seen in abscesses with cultures that were positive for facultative anaerobes only. It is likely that all three of these patients initially had mixed anaerobic and aerobic infections that while being treated with antibiotics yielded only a few anaerobes and/or dead anaerobes in the abscess cavity and thus remained undetectable in cultures.
More recently, Akutsu et al (25) studied five cases of brain abscesses and observed results similar to those in this study with respect to the metabolite patterns seen in the anaerobic and sterile abscesses. It appears that most of the anaerobic abscesses or the mixed anaerobe and aerobe abscesses contained predominantly Suc and/or Ac in addition to the other metabolites.
In this study, we never observed Suc without Ac in any of the cases. There are a number of studies in which Suc was seen only in conjunction with Ac (5,7,24–26). To our knowledge, there is only one case, that reported by Kim et al (8), in which the results were inconsistent with our finding in that signal resonances of Suc were present without resonances of Ac but with resonances of N-acetylaspartate and creatine peaks—findings that suggest brain abscess contamination of the brain parenchyma. Because Ac signals at 1.92 ppm and N-acetylaspartate signals at 2.02 ppm resonate at very close frequencies, it is beyond the spatial resolution capability of 1.5-T clinical MR imaging systems to differentiate these signals at in vivo spectroscopy in the presence of voxel contamination from adjoining brain tissue, and, thus, erroneous assignments may result. Harada et al (22) observed similar erroneous assignments in their study.
The nonspecific spectral pattern of the sterile (ie, group 3) abscesses was probably due to the fact that the patients with these abscesses were undergoing extended treatment with antibiotics (21,26), which probably resulted in the suppression of bacteria inside the abscess cavity and subsequently no bacterial growth in the cultures.
The mechanism for Lac production is known to be similar in anaerobic and aerobic bacteria. It has also been reported that anaerobes reuse the produced Lac for energy production and incorporate it into the tricarboxylic acid cycle at the oxaloacetate level (28). Among the three groups of patients, we observed the lowest concentration of Lac in the group 1 patients at ex vivo spectroscopy (Table 3). This was also seen in the in vivo (Table 2) and ex vivo (Table 4) Lac/AA ratio data: The anaerobic abscesses had lower Lac/AA ratios than the other abscesses (Tables 2 and 4). The variation in Lac/AA ratios may have been influenced by changes in AA or Lac levels. The concentrations of AAs calculated from the ex vivo data did not change greatly among the groups, whereas Lac levels showed wide variations. The Lac/AA ratios at in vivo MR spectroscopy showed significant differences among the groups (P = .008) that were not observed at ex vivo MR spectroscopy. This discrepancy between the in vivo and ex vivo Lac/AA ratio data was probably due to the difference in the quantification of AA peaks between these two studies.
In the in vivo study, we quantified the 0.90-ppm peak contributed by valine, leucine, and isoleucine together. However, in the ex vivo study, only valine and isoleucine signal resonances were quantified and the leucine resonance was excluded. The other probable reason for this inconsistency between the in vivo and ex vivo data was the metabolic changes that took place during the period between pus removal and freezing for the ex vivo study (24). The lowest concentration of Lac in the anaerobic abscesses is evidence that anaerobes reuse the produced Lac for energy production. The quantification of Lac from in vivo and ex vivo MR spectroscopic data may help one understand physiologic features in terms of the possible metabolisms of various microorganisms that inhabit abscesses.
The presence of a higher concentration of Ac than Suc in both the in vivo and ex vivo MR spectroscopic studies indicates that the biochemical reactions leading to Ac production are more favorable than those that lead to Suc production in anaerobes and that Suc sometimes may be produced in concentrations that are beyond the sensitivity of in vivo spectroscopy.
Most brain abscesses are treated by using a combination of medical and surgical therapies. However, abscesses smaller than 2 cm in diameter are more commonly treated with medical therapy alone (2). The antimicrobial regimens commonly recommended for the therapy of brain abscesses are empiric (29), and a combination of antibacterial agents that may be effective against a wide range of microorganisms is tried. The categorization of brain abscesses with in vivo MR spectroscopy may be of value for the rapid detection of anaerobes and thus may help to more appropriately treat patients who are undergoing medical therapy only.
We conclude that it may be possible to differentiate anaerobic brain abscesses from other pyogenic (ie, aerobic or sterile) brain abscesses on the basis of metabolite patterns seen at in vivo proton MR spectroscopy. With the availability of higher-field-strength clinical MR imaging units and two-dimensional localized in vivo spectroscopy, it will be possible to detect the overlapping metabolites that cannot be differentiated with one-dimensional localized spectroscopy and thus to further subcategorize these abscesses.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, R.K.G.; study concepts and design, R.K.G.; literature research, M.G., S.C., R.K.; clinical studies, J.C., M.H.; experimental studies, M.G., S.C.; data acquisition, M.G., S.C., R.K.; data analysis/interpretation, M.G., S.B.R., S.C., R.K.; statistical analysis, M.G., K.N.P., S.B.R.; manuscript preparation, M.G., R.K.G., K.N.P., J.C.; manuscript definition of intellectual content and revision/review, R.K.G., K.N.P., M.K.M., M.H.; manuscript editing, M.G., S.C., R.K., R.K.G.; manuscript final version approval, R.K.G., K.N.P., M.H.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H.-S. Liu, H.-W. Chung, C.-J. Juan, S.-Y. Tsai, C.-Y. Wang, C.-C. Chan, G.-S. Huang, M.-C. Chou, C.-S. Lee, C.-W. Ko, et al. Anomalous J-Modulation Effects on Amino Acids in Clinical 3T MR Spectroscopy AJNR Am. J. Neuroradiol., October 1, 2008; 29(9): 1644 - 1648. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P.H. Lai, H.H. Weng, C.Y. Chen, S.S. Hsu, S. Ding, C.W. Ko, J.H. Fu, H.L. Liang, and K.H. Chen In Vivo Differentiation of Aerobic Brain Abscesses and Necrotic Glioblastomas Multiforme Using Proton MR Spectroscopic Imaging AJNR Am. J. Neuroradiol., September 1, 2008; 29(8): 1511 - 1518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.A. Brunberg and for the Expert Panel on Neurologic Imaging Ataxia AJNR Am. J. Neuroradiol., August 1, 2008; 29(7): 1420 - 1422. [Full Text] [PDF]
|
|
|
|
|
|
G. Luthra, A. Parihar, K. Nath, S. Jaiswal, K.N. Prasad, N. Husain, M. Husain, S. Singh, S. Behari, and R.K. Gupta Comparative Evaluation of Fungal, Tubercular, and Pyogenic Brain Abscesses with Conventional and Diffusion MR Imaging and Proton MR Spectroscopy AJNR Am. J. Neuroradiol., August 1, 2007; 28(7): 1332 - 1338. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 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]
|
|
|
|
|
|
R. K. Gupta, K. M. Hasan, A. M. Mishra, D. Jha, M. Husain, K. N. Prasad, and P. A. Narayana High Fractional Anisotropy in Brain Abscesses versus Other Cystic Intracranial Lesions AJNR Am. J. Neuroradiol., May 1, 2005; 26(5): 1107 - 1114. [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]
|
|
|
|
 |
|
 T. M. Holmes, J. R. Petrella, and J. M. Provenzale Distinction Between Cerebral Abscesses and High-Grade Neoplasms by Dynamic Susceptibility Contrast Perfusion MRI Am. J. Roentgenol., November 1, 2004; 183(5): 1247 - 1252. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
