A number of research studies conducted the analysis of commercially available MALDI-TOF MS systems to evaluate their performance to identify the routinely encountered bacterial isolates in clinical microbiology laboratories (Table X). The vast majority of errors in these reports are attributed to incomplete population of databases, technician error in database assembly and during data acquisition or lack of the MS spectra to differentiate similar species.
If you need assistance with writing your essay, our professional essay writing service is here to help!Essay Writing Service
Seng and co-authors reported a MALDI-TOF MS based study of 1660 bacterial isolates representing 109 different species. Identification was achieved for 84.1% of isolates to the species level but 11.3% at the genus level. They found that nearly 50% of S. pneumoniae isolates were misidentified as Streptococcus parasanguinis as the database included two S. parasanguinis reference spectra and three S. pneumoniae only. In addition, 70% of Stenotrophomonas maltophilia were misidentified as Pseudomonas hibiscicola and all of Shigella sonnei as E. coli. Moreover, 64 isolates were not detected by MALDI-TOF MS. However, these misidentifications and lack of identification for 64 isolates were due to improper database entries (Seng et al., 2009).
In a retrospective study on 1116 clinical isolates, MALDI-TOF MS (Bruker Daltonics) could achieve 1062 (95.2%) overall correct identification at the species level. Correct identification of non-fermenting gram negative rods, streptococci, enterobacteriaceae, staphylococci and enterococci are 79.7%, 93.7%, 95.5%, 99.5% and 100% respectively (Eigner et al., 2009)
Cherkaoui and co-authors first compared the performance of two commercially available MALDI-TOF systems (Bruker Daltonics and Shimadzu) with routine biochemical tests commonly used for bacterial species to identify 720 isolates representing 33 different genera. They found that correct identification at the species level was achieved in 99.1% of cases by Bruker and 88.8% by Shimadzu. Among the species, the percentage of isolates identified was lowest in anaerobes (only 17% by Bruker and 0% by Shimadzu) i.e. they were frequently unidentified. In addition, streptococci were poorly identified only 41% by both MS devices (Cherkaoui et al., 2010).
In a study of 1371 clinical isolates, it was found that 1278 (93.2%) could be identified at the species level and 73 (5.3%) at the genus level by MALDI-TOF MS (Bruker Daltonics) (Bizzini et al., 2010).
Similar results were published by analyzing 980 microbial isolates in which 61 yeast isolates are also included. Overall correct MALDI-TOF based identification at the species level was 92%. Discrepancies between biochemical identification and MALDI-TOF was verified by using 16S rRNA gene sequencing and correct species identification by using MALDI-TOF MS was observed in 97.7% of Enterobacteriaceae, 92% of non-fermentative Gram-negative bacteria, 94.3% of staphylococci, 84.8% of streptococci, and 85.2% of yeasts. Misidentifications are related with a lack of reference spectra for rare species in database and viridans streptococci and pneumococci are frequently misidentified(van Veen et al., 2010)
Recently, a study compared the performance of three MALDI-TOS MS systems Microflex LT (Bruker Daltonics), Vitek MS RUO (Axima Assurance-Saramis database; bioMérieux) and Vitek MS IVD (bioMérieux). Total 1129 isolates were tested on the Microflex LT and Vitek MS devices and spectra were analyzed by three databases namely Biotyper (Bruker Daltonics), Saramis, and VitekMS(bioMérieux). These databases performed almost equally and 93% of isolates were accurately identified to the species level (Martiny et al., 2012).
Regarding to the delay of MALDI-TOF MS bacterial identification, it was estimated that the average delay for the identification of a bacterial isolate is 6 minutes compared to the conventional gram staining methods that will take 5 to 48 hours to get the result. Specifically, for MALDI-TOF MS identification (15 isolates; 4 spots per isolate), it took 25, 15 and 50 minutes for plate preparation, plated loading and plate reading and spectra interpretation respectively resulting in a mean delay of 6 minutes an isolate. (Seng et al., 2009). Further studies will be necessary to identify the impact of speed of MALDI-TOF MS on the patient treatment outcomes in various settings including rapid identification of bacterial pathogens and their antibiotic susceptibility testing.
Despite higher initial cost of the instrument, MALDI-TOF MS technology is more cost-effective than the conventional phenotypic identification methods. In addition, it is also user-friendly requiring the low to medium level of training for laboratory workers whereas the conventional methods of bacterial identification such as gram staining and Vitek system require relatively higher level of training and more processing time. Total cost of MALDI-TOF MS identification was much less than that of conventional identification (€2.44 Vs € 4.60-13.85) as seen in Figure XXX. In other words, it cost only 22%-32% of the total cost of conventional phenotypic identification. Cost calculation was based on addition of the cost of specific consumables, salary for employees and 5-year depreciation of the apparatuses such as apparatus used in gram staining, microscope, identification apparatus and mass spectrometer (Seng et al., 2009). Therefore, the reduction of cost is one of the advantages of MALDI-TOF MS.
A similar cost-estimation in US dollars was carried out by Cherkaoui and co-authors. However, the results represented the cost to the laboratory rather than the patients. Mass spectrometry instruments are costly and the expenses are comparable to 16S sequencing devices and automated blood culture devices. Nevertheless, the marginal estimated cost of MS identification for an isolate was US$ 0.50. On the other hand, it cost US$ 10 per isolate for phenotypic identification using automated devices (Cherkaoui et al., 2010). Therefore, application of MALDI-TOF MS is more economical than the other methods for bacterial diagnostics. This will have a large impact on the healthcare system in terms of saving cost and time and achieving fast and reliable bacterial diagnosis for a better management of infectious diseases.
For the diagnosis of UTIs, blood and bacteria present in the urine hamper the urinary proteomic analysis by changing peptide mass signals (Fiedler et al., 2007). In addition, the presence of bacterial overgrowth in the urine sample could impair the proteomic analysis and thus it is recommended to centrifuge to remove epithelial cells and leukocytes in pellet and store the sample at 4oC or add boric acid or NaN3 to avoid bacterial overgrowth (Thongboonkerd and Saetun, 2007).
MALDI-TOF combined together with UF-1000i urine flow cytometry is useful for the fast, accurate and direct analysis of UTIs (Wang et al., 2013). Wang and co-authors compared conventional identification, MALDI-TOF MS and urine flow cytometry. In this study, the urine flow cytometry, as a prescreening step, eliminates negative samples rendering downstream utilization of MALDI-TOF MS more efficient whereas MALDI-TOF MS is used to determine the presence of bacteria in remaining positive samples. The results are compared with those from phenotypic methods and discrepancies are resolved by 16S rRNA gene sequencing. 943 (64%) of 1456 samples from patients showing UTI symptoms are negative by using these three testing methods. The combined method is consistent with the conventional method in 1373 of total 1456 cases (94.3%), and gives the correct result in 1381 of 1456 cases (94.8%).
The identification process could be hastened by direct identification of microbes from blood culture bottle broth using MALDI-TOF systems. The blood culture broth should be processed in order to limit disturbances from hemoglobin and blood cells and to concentrate the bacterial pathogens in the blood. The procedure involves centrifugations steps to remove red blood cells and then lysis by formic acid or trifluoroacetic acid. Staphylococci identification can be improved by using formic acid instead of trifluoroacetic acid in the workflow and the overall correct identification rose from 59% to 76% (La Scola and Raoult, 2009).
Our academic experts are ready and waiting to assist with any writing project you may have. From simple essay plans, through to full dissertations, you can guarantee we have a service perfectly matched to your needs.View our services
There are some reports concerning with the possibility of direct bacterial identification in blood culture samples. Organisms are detected by automated blood culture systems and MALDI-TOF MS devices identify the detected organisms in these studies (La Scola and Raoult, 2009, Prod’hom et al., 2010, Stevenson et al., 2010). Using various protocols, the processing of samples can be divided into four steps (1) collection of specimen (2) removal of RBCs (3) protein precipitation and finally (4) protein extraction and solubilization. After that, MALDI-TOF MS will analyze the samples. Protocols using extraction are better and more effective than those with the intact cell method (La Scola and Raoult, 2009, Ferreira et al., 2011).
Performance in detecting bacteria in positive blood culture broth was tested by a number of studies. In these studies, they indicate that MALDI-TOF MS software was unable to reliably detect all organisms in mixed cultures. Firstly, La Scola and Raoult tested 584 positive blood cultures including 562 specimens containing a single bacterial species and the remaining specimens containing mixed bacterial species. Two different protocols were used. They could detect more gram-negative bacteria than gram-positive bacteria; 94% of gram-negative organisms were identified compared to 67% of gram-positive organisms. Moreover, Viridans streptococci could not be identified at all. In the 22 positive blood culture broths with mixed species, only one of the species was detected in 18 samples and then 2 samples could not identified at all and 2 samples as false species. It was recommended to use gram staining to achieve optimal detection of mixed species (La Scola and Raoult, 2009).
In a further study by Prod’hom et al., correct identification was obtained for 78.7% of 122 positive blood cultures. Identification was failed mainly for streptococci and staphylococci. 8 out of 10 Streptococcus pneumoniae were not identified and the two others were identified only with a low score. In addition, detection of encapsulated microorganisms such as Haemophilus influenzae and Klebsiella pneumoniae was not accurate (Prod’hom et al., 2010).
A third study by Stevenson et al. used the BioTyper software version 2.0 to identify 212 positive cultures representing 60 species and 32 genera. Most commonly due to inadequate number of bacteria in blood culture broth, less than 42 (19.8%) of isolates with spectral score of <1.7 could not identified. 162 (95.3%) of isolates were accurately identified at the genus level with scores of ≥1.7 and at the species level, 138 (65%) were obtained for correct identification with scores of ≥1.9. All of eight Streptococcus mitis resulted in misidentification as Streptococcus pneumoniae (Stevenson et al., 2010). All organisms misidentified as being S. pneumoniae by MALDI-TOF MS are recommended to confirm by doing a bile solubility test directly on the blood culture broth(Murray, 1979).
A study with larger sample size of 277 aerobic and anaerobic isolates achieved the accurate identification for 95% at the species level. Mismatching was mainly due to insufficient bacterial numbers and occurred with gram positive samples (Christner et al., 2010).
Recently, Chen and co-workers compared the two different commercial MS systems namely Vitek MS IVD (bioMérieux) and the Microflex LT Biotyper (Bruker Daltonics) for the microbial identification directly from 202 blood culture-positive specimens. 181 of them were monomicrobial and the remaining 21 blood cultures were polymicrobial. Sample processing was done using the Bruker Sepsityper kit. Evaluation of performance of these devices revealed that the Biotyper system produced more correct identifications than the Vitek MS IVD system (177 Vs 167 out of 181 monomicrobial specimens). Both systems performed genus to species level identification for more than 90% of specimens (Biotyper, 97.8%; Vitek MS IVD, 92.3%). In polymicrobial blood cultures, Bruker Biotyper generated polymicrobial identifications in 5 out of 21 mixed-culture specimens (23.8%). So far, both systems are not reliably ready yet for direct use with polymicrobial cultures (Chen et al., 2013).
In parallel with bacterial identification, rapid testing of antibiotics sensitivity of bacteria is also critical for the timely implementation of targeted antibiotics treatment to patients. Although there have been rapid testing kits for routinely encountered resistant strains such as MRSA, these are limited to be applicable to a few number of bacterial species. Moreover, automated phenotypic identification systems of bacteria e.g. the Vitek-2 and the Phoenix systems can also be used to identify antimicrobial susceptibilities but these methods will take more time to get results. Therefore, MALDI-TOF MS may be adopted to serve as a universal platform for not only rapid identification but also covering to a wide range of resistant bacterial species in the workflow of the microbiological laboratory.
Camara and Hays published a study which focused on a specific beta-lactamase peak for ampicillin-resistant Escherichia coli which were grown in Luria-Bertani broth mixed with penicillin. From the protein extracted and spotted onto a MALDI-TOF plate, beta-lactamase peak of approximately 29 kDa was successfully found in the spectra. By using whole cells for bacterial identification or using the other matrices instead of sinapinic acid, the peak of beta-lactamase was not detectable. The 29kDa was then identified by SDS-PAGE and LC-MS peptide mass fingerprinting as beta-lactamase(Camara and Hays, 2007).
Recently, Schaumann and co-workers tried to evaluate how MALDI-TOF MS can discriminate among ESBL-producing or MBL-producing and non-producing E.coli, Klebsiella pneumoniae and Pseudomonas aeruginosa. The correct classification rates achieved showed that strains producing ESBLs and MBLs tend to produce different spectral patterns compared to the nonproducing counterparts, but 70% accuracy was not yet reliable enough for routine diagnostics (Schaumann et al., 2012).
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this essay and no longer wish to have your work published on UKEssays.com then please: