 |
| About Bioline | All Journals | Testimonials | Membership | News |
|
||||||
|
||||||
Indian Journal of Medical Microbiology, Vol. 28, No. 4, October-December, 2010, pp. 348-353 Original Article Utility of in vitro proton magnetic resonance spectroscopy in aetiological characterisation of brain abscesses S Menon1, R Bharadwaj1, AS Chowdhary1, DV Kaundinya1, DA Palande2 1 Department of Microbiology, Grant Medical College & Sir J.J. Hospital, Mumbai - 400 008, India Date of Submission: 27-May-2010 Code Number: mb10103 PMID: 20966567 Abstract Purpose: Brain abscesses often present an aetiological dilemma. Microscopy is insensitive and culture techniques are time consuming. Hence, a new rapid technique in vitro Proton Magnetic Resonance Spectroscopy ( 1 HMRS) was evaluated for its usefulness in the identification of aetiology of brain abscesses. Keywords: In vitro proton MR spectroscopy, pyogenic brain abscesses, tuberculous brain abscess Introduction Brain abscess is a focal, intracerebral infection that begins as a localised area of cerebritis and develops into a collection of pus surrounded by a well-vascularised capsule. [1] Currently, they are usually diagnosed on the basis of clinical manifestations, Computed tomography (CT) and Magnetic Resonance Imaging (MRI). However, these techniques do not assist in an aetiological diagnosis of a brain abscess. Conventionally, the aetiological diagnosis is only possible after microbiological culture of drained pus from brain abscesses. These culture techniques are time consuming and result in delayed patient management. Making a rapid aetiological diagnosis of a brain abscess is very essential in order to provide the appropriate therapy and thus a decrease in the morbidity and mortality. This would rationalise the management of brain abscess and considerably decrease the mortality associated with it. Magnetic resonance (MR) spectroscopy has shown the potential of identifying micro-organisms from infections. [2],[3] The organisms are subjected to an intense magnetic field. The differences in the cell wall constituents, cytoplasmic constituents and metabolites give rise to characteristic signals when subjected to this magnetic field. In fact, attempts have also been made to use this technique for in vivo diagnosis of infecting agents. [4],[5],[6],[7] In the present study the in vitro Proton Magnetic Resonance Spectroscopy ( 1 HMRS) was evaluated for its role in the diagnosis of brain abscesses. Materials and Methods Aspirated pus specimens from 39 patients with brain abscesses were collected during neurosurgery, either by burr hole or by craniotomy, and sent to Department of Microbiology for the subsequent workup for specific diagnosis after obtaining proper consent. This work was carried out after obtaining necessary ethical clearance from the institutional ethical committee. These pus specimens were further subjected to conventional microbiological analysis and in vitro 1 HMRS for aetiological identification. Conventional microbiological analysis The pus was collected in plain sterile container for aerobic culture, and the pus for anaerobic culture was collected in Robertson's cooked meat medium (RCM). Two direct smears were made and subjected to Gram and Ziehl-Neelsen stain. Pus was inoculated on Blood agar and Mac'Conkeys agar and incubated aerobically at 37°C for 24 hours. Similarly, pus sample was also inoculated on Chocolate agar at 37°C in 5-10% CO 2 and incubated for 48 hours. All the isolates obtained were identified by conventional techniques. [8] Pus from brain abscess was also inoculated at the bedside whenever possible or inoculated in the RCM for anaerobes. It was then inoculated on Blood agar, Neomycin Blood agar and Bacteroides Bile Esculin Agar. All the anaerobic culture plates were incubated at 37°C for 48 hours-7 days in a Gas-Pak system (Dynox/Dynamicro, Mumbai, India). These plates were observed initially after 48 hours for growth. The plates showing no growth were further incubated for 7 days before giving a final negative report. A single colony of each distinct type was checked for aerotolerance in CO 2. Aerotolerant colonies were disregarded. Only strict anaerobes were further followed. Preliminary grouping of isolated anaerobes was done using Kanamycin 1 mg, Colistin 10 μg, Vancomycin 5 μg and Sodium polyanethol sulfonate 1 mg (Becton-Dickinson, NJ, USA). Identification of anaerobes was further confirmed by carbohydrate fermentation, gelatin liquefaction, bile resistance, indole production and lipase production. [8] Pus specimens showing the presence of acid fast bacilli on Ziehl-Neelsen stain were further inoculated on Lowenstein-Jensen medium and incubated at 37°C for 8-12 weeks. Colonies identified were confirmed to be acid fast by Ziehl-Neelsen stain and then identified by conventional biochemicals. [8] When Gram stain was suggestive of fungal aetiology, KOH mount was also done. Fungal cultures were set up on Sabouraud's Dextrose Agar (SDA) with and without antibiotics. Identification of fungus was done as per the conventional methods. [9] In vitro proton magnetic resonance spectroscopy All these 39 pus specimens were also subjected to in vitro 1 HMRS using Mercury plus Varian 300 MHz (7.05 T) NMR spectrometer. One hundred microlitre of pus specimen was loaded in 5 mm NMR tube and deuterium oxide (D 2 O Armar Chemicals Dottingen Switzerland) was added to make the approximate volume 0.6 ml and then subjected to NMR analysis. After the NMR analysis (256 scans), the different peaks obtained were noted. Referencing was done with the water peak (4.8 ppm). The interpretation of peaks was done according to the available literature. [2],[7] The spectra thus generated were then correlated with culture results. Results A total of 39 pus specimens collected from brain abscesses during neurosurgery were subjected to conventional microbiological culture and in vitro 1 HMRS. Out of 39 specimens, 26 showed the growth of aerobic or anaerobic bacteria. Two abscesses were due to Mycobacterium tuberculosis. A solitary case each of Actinomycotic brain abscess and Cladophialophora bantiana also were encountered. Remaining nine pus specimens did not show any growth after conventional culture [Table - 1]. MR spectra of all these specimens were studied and these were analysed with respect to the culture results. Four characteristic patterns emerged [Table - 2] and [Figure - 1]a-d. The analysis revealed that spectra were characteristic to the nature of the infecting agents. Further, the detailed analysis of spectra was done with respect to individual organisms. The constant and variable peaks were noted as shown in [Table - 3]. There was consistency in the spectral patterns obtained in the pus infected by the same organism. The constant peaks were identified for individual organism. Discussion Pyogenic infections of the brain are usually associated with a high rate of morbidity and mortality. They pose a serious diagnostic dilemma and a critical therapeutic challenge. It is necessary to characterise the pathogens causing the abscess in order to institute early appropriate therapy. CT and MRI are the two main noninvasive modalities used in the diagnosis of brain abscess. These techniques do not help in the aetiological diagnosis. Aetiological diagnosis is only possible after draining the pus from the abscess, which is further subjected to conventional microbiological culture. Conventional microbiological culture techniques are time consuming. Hence a new technique, "in vitro 1 HMRS", was evaluated for its diagnostic utility in the rapid diagnosis of brain abscesses in the present study. All the pus specimens which grew pyogenic organisms on culture showed the presence of multiplet of amino acid-lipid at 0.9 ppm, while this peak was conspicuously absent with tuberculous brain abscesses. Pyogenic brain abscesses contain large amounts of neutrophils and proteins. After the breakdown of neutrophils, they release proteases which in turn degrade the protein into amino acids [10] and this could be the reason of a peak of amino acids at 0.9 ppm seen with the pyogenic brain abscesses, while tuberculous brain abscesses are loaded with Mycobacteria along with lymphocytes and a very small number of neutrophils in the pus. [11] Tuberculous abscesses have abundant lipid as Mycobacteria are rich in lipid; however, they lack proteolytic enzymes. [12] The absence of amino acids at 0.9 ppm in tuberculous brain abscesses may be probably due to the presence of large amounts of Mycobacteria and the lack of proteolytic enzymes resulting in poor breakdown of proteins into amino acids. Gupta et al.[5] and Husain et al.[12] also had reported similar observations in their study of the tuberculous brain abscesses. Gupta et al.[5] in their study also attempted to determine whether pyogenic brain abscesses could be differentiated from tuberculous brain abscesses by use of in vivo MR spectroscopy and magnetization transfer (MT) MR imaging. In their study, all the 27 patients with pyogenic brain abscesses showed the presence of multiplet of amino acids at 0.9 ppm, while the three patients with tuberculous brain abscesses did not show any amino acid at 0.9 ppm. Grand et al.[13] also assessed the usefulness of the 0.9 ppm peak from amino acids (CH 3 moieties from valine, leucine and isoleucine) for the differentiation of brain abscess and tumors using in vivo Proton MR Spectroscopy. Thirty-four patients with cystic intracerebral mass lesions (28 tumors, six abscesses) were examined by 1 H MR spectroscopy in vivo. The proton spectra revealed amino acids only in the abscesses. Detection of amino acids at 0.9 ppm in brain abscesses applies to only pyogenic infection. In the case of nonbacterial infections, no amino acids were detected. Eight pus specimens which showed obligate anaerobes on culture revealed the presence of lactate-lipid at 1.3 ppm (100%), acetate at 1.92 ppm (100%), succinate at 2.4 ppm (75%) along with the multiplet at 0.9 ppm. Lai et al.[10] observed the presence of lactate, cytosolic amino acids and acetate in all six patients who showed the presence of obligate anaerobes or a mixture of obligate and facultative anaerobes, while succinate was observed in four patients. They found that succinate was always present in conjunction with acetate but never present alone in any of their patients. Even we noticed a similar finding in the present study. Increase in lactate, acetate and succinate presumably originated from the enhanced glycolysis and fermentation of the infecting organism. [4],[13] Succinate is known to be a marker for anaerobes. [14] The present study also found the presence of succinate at 2.4 ppm in 75% of the pus specimens which yielded anaerobes. The constant and the variable peak analysis helped us to create a diagnostic algorithm for anaerobes. [15] The 16 pus specimens, which showed the presence of facultative anaerobes on culture, revealed lactate acetate along with the multiplet at 0.9 ppm. Pus specimens which yielded obligate aerobes like Pseudomonas aeruginosa on culture showed the presence of only lactate along with the multiplet at 0.9 ppm. Pus specimen which grew Cladophialophora bantiana on culture showed a triplet of ethanol at 1.16 and a singlet at 1.18 ppm. The other minor peaks which showed their presence was glycine at 3.5 ppm and multiple peaks at 3.6 and 3.8 ppm (trehalose). The multiplet of amino acids-lipid at 0.9 ppm was also found. Bacterial abscesses showed the presence of acetate (1.9 ppm) and succinate (2.4 ppm) which were absent in tuberculous as well as Cladophialophora brain abscess pus [Figure - 1]e. Actinomycotic brain abscess spectra were conspicuously different from all other spectra and showed the presence of two intense peaks at 0.9-1.3 ppm region [Figure - 1]f. There are very few studies describing the MR spectra of fungi. As that of bacterial brain abscesses, the metabolites which appear with fungi are totally different. Ethanol, a product of glucose fermentation in yeasts, was reported to be present in the cerebrospinal fluid of a patient with cryptococcal meningitis. [16] Ex vivo Proton Magnetic Resonance Spectroscopy has shown a high concentration of alpha-trehalose and is considered as diagnostic of Cryptococcus neoformans. [17] Siegal et al.[18] reported the possible differentiation of cerebral mucormycosis from bacterial abscesses by the use of MRS. Accordingly, MRS was showing markedly elevated lactate, depleted N-acetyl aspartate and metabolite resonance attributable to succinate and acetate in that case, which was essentially similar to that of bacterial abscess, but without the commonly seen resonances of the amino acids valine, leucine and isoleucine. PMRS in an isolated case has shown elevation of glutamine-glutamate (2.1 ppm), lactate (1.3 ppm) and amino acids (0.9 ppm) at the central non-enhancing part of the Aspergillus abscess. [19] MR spectroscopy findings in two other clinically important pathogenic yeasts, Candida albicans and Saccharomyces cerevisiae, and the fungus, Aspergillus fumigatus, were investigated and found to be different from those of C. neoformans. The 1D and 2D COSY spectra of C. albicans and S. cerevisiae revealed dominant lipids, whereas those of A. fumigatus were characterised by resonances from amino acid residues and carbohydrates. [17] Our extensive literature survey did not yield any report of MR spectroscopic finding (in vitro or in vivo) of Cladophialophora. bantiana as well as also that of Actinomycetes. In nine pus specimens, conventional culture remained sterile. This may be because the samples were collected from tertiary care centre and patients had been treated with long duration of broad spectrum of antibiotics before they reported to hospital. Other possibility is the presence of obligate anaerobes which might have lost during transit. Pus from culture negative abscesses also showed some spectra. An attempt was made to analyse these spectra. These spectra were then compared with the spectral database generated in this study after analysing 28 culture positive pus specimens. The aetiologic agent of culture negative abscesses was postulated depending on the spectral patterns. The aetiological agent could be identified in all the nine culture negative abscesses. The infecting agent could be identified up to the species level in seven (77%) cases based on the database generated in the study [Table - 4]. Garg et al. [6] reported that the sterile cultures in their study were characterised by the presence of lactate along with the amino acids. Lai et al.[7] found spectra of two sterile brain abscesses as lactate, acetate with or without succinate and lactate with amino acids. However, they have not postulated the possible organism depending on the spectra. Thus, our experience with in vitro 1 HMRS was encouraging. It could be used to diagnose a pyogenic brain abscess and possibly to differentiate anaerobic from aerobic brain abscess on the basis of spectral patterns observed. In vitro 1 HMRS has shown the potential to assist in the rapid aetiological identification of brain abscesses once the pus is drained from the abscesses, and thus, the targeted therapy can be started at the earliest. Further, we would like to hypothesise that the spectral database generated in the in vitro study also can be used for the interpretation of in vivo MR spectroscopy data. This in turn may help us to identify the causative agent without draining the pus. References
Copyright 2010 - Indian Journal of Medical Microbiology The following images related to this document are available:Photo images[mb10103t4.jpg] [mb10103t2.jpg] [mb10103t1.jpg] [mb10103t3.jpg] [mb10103f1.jpg] |
| |||||||||
![[Table - 1]](/showimage?mb/photo/mb10103t1.jpg){kind=link}
![[Table - 2]](/showimage?mb/photo/mb10103t2.jpg){kind=link}
![[Figure - 1]](/showimage?mb/photo/mb10103f1.jpg){kind=link}
![[Table - 3]](/showimage?mb/photo/mb10103t3.jpg){kind=link}
![[Table - 4]](/showimage?mb/photo/mb10103t4.jpg){kind=link}