Bacteriology

Specimen collection, transport, and stability

Upon arrival in the laboratory, information about patient specimens (e.g., specimen type, source, collection date/time) is recorded in the laboratory information system. Depending on the physician’s order and the type of specimen, wet mounts and smears can be prepared and stained for microscopic examination; the findings of these direct microscopic examinations can be immediately released and reported. Often, these preliminary results are used to establish a presumptive diagnosis and initiation of empiric therapy. Moreover, particularly with respect to direct Gram stains, results can help to assess specimen quality and provide guidance for interpreting subsequent culture results. For a summary of key microbiology stains, see Table 84.1 and Fig. 84.1 , and for common Gram stain results and example organisms associated with each, see Table 84.2 and Fig. 84.2 .

Culture media

After receipt of the specimen into the laboratory information system, specimens are inoculated to appropriate primary plating media. Primary plating media are the nutritive media that a clinical specimen is first inoculated to in order to recover common bacteria that might be present. These media can be categorized as general purpose, selective (supporting the growth of one group of bacteria over another), enriched (contain particular nutrients that support the growth of fastidious bacteria), and differential (allow for phenotypic differentiation of bacteria via sugar utilization, hemolysis, etc.). These media categories are not mutually exclusive. Some common examples of primary media include 5% sheep’s blood agar (BAP; general purpose and differential—hemolysis), chocolate agar (CHOC; general purpose and enriched), and MacConkey agar (MAC; selective and differential—gram-negative organisms and lactose fermentation). The primary plating media used for any clinical specimen is based on the physician’s order in conjunction with the anatomic site from which the specimen was obtained. Certain specimens may require additional processing prior to culture inoculation. For example, cerebrospinal fluid (CSF) and other body fluids can be centrifuged or otherwise concentrated if sufficient volume is available, as bacteria isolated from this source are typically found in low numbers. Furthermore, decontamination methods may be required to assure that slowly growing pathogens, such as Mycobacterium spp., are not outcompeted by normal microbiota (i.e., contaminants) when specimens are collected from nonsterile sites. Once inoculated, plates are incubated under conditions that support the growth of potential bacterial pathogens suspected from the particular anatomic site of the specimen. Typically, these conditions include incubation at 35 ± 2 °C, in either room air or 5% CO ₂ .

Once placed into major organism groups, based on Gram stain characteristics and growth on media, an approach to identify the bacteria can be delineated. There are numerous approaches to the identification of bacteria which are recovered from clinical specimens using routine culture media. As previously mentioned, the main media used routinely for most clinical specimens submitted for routine bacterial culture are BAP, CHOC, and MAC. In addition, a large number of primary plating media are used by the clinical microbiology laboratory for certain patient populations and specimen types to recover potential bacterial pathogens. For example, Burkholderia cepacia selective agar (BCSA) or similar can recover B. cepacia group isolates from respiratory specimens obtained from patients living with cystic fibrosis (CF). Another example is the specific recovery of Salmonella or Shigella species from stool with xylose-lysine-dextrose (XLD), Hektoen Enteric (HE), or Salmonella-Shigella (SS) agars. If a specimen is submitted in which anaerobic bacteria may play an important role in a disease process, anaerobic primary plating media are also inoculated and incubated under anaerobic conditions. These media include Brucella blood agar (BBA), Brucella laked blood agar with kanamycin and vancomycin (LKV), Bacteroides bile esculin agar (BBE), and phenylethyl alcohol agar (PEA). Table 84.3 and Fig. 84.3 compiles both nutritive and differential primary plating media. The potential pathogens and their recovery by bacterial culture are discussed in their respective clinical syndromes later in this chapter.

Some bacteria can be recovered by culture but require special media and/or growth conditions, while others cannot be cultured at all. Detecting organisms that are difficult to grow relies on the prior request of the clinician for specific bacteria or groups of bacteria. In other words, some bacteria will not be detected using routine primary plating media, and their presence can be detected only through use of special selective media, special incubation conditions, serologic methods, or direct molecular detection methods ( Fig. 84.4 ).

Laboratory automation

In recent years, laboratory automation has been introduced into some clinical microbiology laboratories. Whereas each of the available laboratory automation systems can be used for the majority of the culture workflow (processing, plating, incubation, and reading), many laboratories use only the processing modules to automate just the processing and plating of specimens onto primary culture media. The two most commonly used systems are the Walk-Away Specimen Processor (WASP) by Copan Diagnostics, Inc. (Murrieta, CA), and the Kiestra InoqulA by Becton, Dickinson and Company (Franklin Lakes, NJ). The WASP system uses disposable calibrated loops, whereas the Kiestra InoqulA system uses a pipette to dispense a known volume, followed by streaking of the plate via magnetic beads. Both systems are able to vortex specimens, de-cap/re-cap specimen containers, and streak plates according to a number of customizable patterns. As not all specimens are liquid, both systems allow for some manual specimen preprocessing, and even preinoculation of media.

A huge benefit of using laboratory automation systems for incubation of cultures is the ability of these systems to image each individual plate of an associated culture while never removing those plates from their incubation conditions. This allows for faster growth and more robust yield of microorganisms from cultured specimens, especially for specimens that may contain fastidious organisms. For additional information on automation in the clinical laboratory, refer to Chapter 29 .

Molecular testing

Nonphenotypic methods, such as target-specific nucleic acid amplification or DNA sequencing, require different preanalytical considerations than those typically observed in the routine clinical microbiology laboratory. One of the greatest advantages of molecular testing is its ability to detect low numbers of target-specific DNA or RNA molecules, but this enhanced sensitivity can also contribute to false-positive results if steps to avoid contamination are not enacted. To minimize contamination, molecular testing has historically required dedicated laboratory space that is divided into two to three rooms, where testing proceeds in a unidirectional workflow. The first area or room is dedicated to reagent preparation and must have positive air pressure to prevent the entry of aerosols. The next area should be devoted to sample preparation and be under negative air pressure to retain aerosols that may be generated when working with the specimen. Laboratories with limited space may combine these two areas into one, but doing so will require barrier containment, such as a dead air box or biological safety cabinet, for reagent preparation. The final area in the unidirectional workflow is the amplification detection room. This area is considered “dirty” and should be under negative pressure to minimize amplicon contamination.

Another consideration, beyond the physical space, is the use of dedicated instrumentation and testing materials, such as pipettes, tubes, and vortexes. Consumables or equipment should never be moved to another area without being decontaminated first. The movement of testing personnel between rooms should also be limited to minimize aerosolization and the spread of amplicons. If staff need to move between rooms, there should be dedicated personal protective equipment (PPE) in each area, sticky mats at the entrance of each room, and gloves should be changed periodically, especially in between specimens. Work areas and equipment should also be cleaned at the beginning and end of each day and between each test.

Conventional identification systems

Growth-based organism identification systems evaluate biochemical and enzymatic reactions for an organism of interest. These reactions yield a series of positive and negative results, which are then compared to a database of reactions for known organisms that can then be used to produce an identification. There are several commonly used growth-based systems for organism identification with varying workflows and performance characteristics.

Despite being based on foundational principles in microbiology, these systems all suffer from the same limitation, namely, that they require an organism to be growing well and in the same way as those organisms that were used to generate the reference database. A multitude of variables (antibiotic exposure, the impact of immune system response, biofilm growth, and heterogeneous morphologies) can impact the growth of an individual patient isolate and therefore system performance. With that said, these systems generally work very well for commonly isolated, nonfastidious pathogens, but have difficulty differentiating closely related organisms, uncommonly isolated organisms, and organisms that are metabolically inert.

Automated versus manual phenotypic identification systems

Phenotypic organism identification can be performed using either automated or manual systems. The automated systems require that the user inoculate an identification card or plate, which is placed in an instrument that incubates, analyzes, and interprets the biochemical reactions to produce an identification. The manual systems rely on many of the same principles, but are manually incubated, read, and interpreted. Historically, most laboratories have relied on both methods, with the automated systems serving as the primary identification platform and the manual method serving as the backup or to aid in discrepant resolution.

Because there are numerous manufacturers of both automated and manual identification systems, a comprehensive review of their performance characteristics is beyond the scope of this chapter. A few commonly used automated identification systems include the bioMérieux VITEK 2, the BD Phoenix, and the Beckman Coulter MicroScan. Among manual identification systems, the most common are the series of bioMérieux API identification panels and the Beckman Coulter MicroScan identification panel. The following is a brief overview of some of the key general reactions that are used to generate organism identifications using these systems.

Carbohydrate utilization

Carbohydrate utilization is the backbone of both automated and manual phenotypic identification systems. These reactions are growth-based and require up to a 24-hour incubation in order to register the reaction. The reactions utilize a pH-based indicator that results in a color change. Carbohydrate utilization results in an acidic condition, while protein utilization results in an alkaline condition, both of which are indicated by color changes that are either read automatically or with the naked eye, depending on the system being used.

Enzymatic profile

Enzymatic profiles can be assessed much more quickly than other reactions because they rely on the presence of preformed enzymes, which is not a growth-requiring reaction. The reactions can be read within 6 hours and are indicated by a color change that results from the hydrolysis of a colorless complex.

Visual observation of growth

Some manual systems utilize the simple observation of growth to determine a reaction. These systems are not conducive to automation as the reaction requires observation with the naked eye. Growth-based reactions assess the dependence of an organism on a variety of different substrates and can be read within 24 hours.

Principles of bacterial identification by MALDI-TOF MS

The first step in identifying an organism by MALDI-TOF MS is harvesting at least 10 ⁵ colony forming units (CFUs) of pure (single, isolated, well-sized colony), cultured organism. A single colony is usually selected from solid agar growth and applied to a target plate with a toothpick or a plastic loop ( Fig. 84.5 ). A matrix of α-cyano-4-hydroxycinnamic acid (usually 1 μL) is applied to the colony using a calibrated pipette and allowed to dry. This matrix is critical to the performance of MALDI-TOF MS, because it absorbs laser energy and facilitates the soft ionization process of proteins prior to their migration through the flight tube. An alternative procedure includes adding formic acid (70 to 100%) to the colony prior to the addition of the α-cyano-4-hydrocinnamic acid as an “on-plate extraction.” This procedure has been shown to improve identification rates for some organisms, most commonly gram-positive organisms. For the Bruker platform, if the routine workflow outlined above fails to produce an identification, a “tube extraction” may be performed. In this tube extraction, an organism suspension is first made in approximately 75% ethanol, pelleted, and resuspended in 70% formic acid and an equal volume of acetonitrile. This suspension is centrifuged leaving protein in the supernatant, which can then be applied to the target plate for analysis by the MALDI-TOF MS for organism identification.

Once the organism or protein suspensions have been applied to the target plate, the target plate is entered into the mass spectrometer and analyzed. Through this process, organism proteins are ionized and migrate through the vacuum chamber. The rate at which they migrate is measured (time-of-flight) and is a product of both mass and charge (mass to charge ratio; m/z) . A detector at the end of the MALDI-TOF MS flight tube measures the ionized proteins and generates a spectral profile or “protein fingerprint.” This fingerprint is unique to the organism and can be compared with a database of previously defined spectra of known organisms (see Chapter 22 for more detail). The algorithm for establishing a match between the spectra of the unknown organism and the database spectra differs by manufacturer. After this comparison, an organism name is reported. Although the instrumentation plays an important role in performance, a high-quality database is the most critical factor for reliably identifying an organism.

In contrast to growth-based identification methods, MALDI-TOF MS identifies organisms by analyzing protein fragments between 2000 and 20,000 Daltons. The protein profiles generated are less susceptible to environmental conditions that can confound biochemical identification. Further, in order for biochemical reactions to produce reliable patterns that accurately predict identity, the organism must be grown in a reproducible and robust manner. This is particularly problematic for fastidious and/or anaerobic organisms. In contrast, MALDI-TOF MS is less affected by these constraints to produce a reliable identification. Studies have shown that MALDI-TOF MS reliably identifies organisms grown on a variety of media.

One of the strengths of MALDI-TOF MS is that it provides high-confidence identification results, with some known exceptions. This characteristic is particularly useful for the identification of rarely encountered or fastidious organisms. In contrast, biochemical systems not infrequently generate confident but erroneous results, requiring significant experience to know when to trust a result. It is important to note that MALDI-TOF MS does have difficulty distinguishing some species of bacteria from one another. For these key organisms, laboratories must implement confirmatory testing to ensure that accurate identifications are generated. Table 84.5 for a list of select organisms that may be misidentified by MALDI-TOF MS along with biochemical reactions that may be used to confirm organism identification.

Gram-positive cocci

Though there are many genera of gram-positive cocci, the most commonly encountered genera in clinical specimens are Staphylococcus , Streptococcus, and Enterococcus ( Fig. 84.6 ). MALDI-TOF MS is a reliable method for the identification of both staphylococci and enterococci. Studies have shown that MALDI-TOF MS accurately identifies more than 90% of these organisms, while incorrect identifications are very rare. ^,

Some streptococci are more challenging for MALDI-TOF MS to identify accurately. A well-known limitation for Bruker Biotyper systems is the inability to discriminate between Streptococcus pneumoniae and the S. mitis/oralis group of organisms. As a result, laboratories using a MALDI-TOF MS system that is unable to differentiate these species must rely on biochemical reactions to supplement their testing. Isolates identified by MALDI-TOF MS as either of these two organisms should be tested with optochin and/or bile solubility.

The ß-hemolytic streptococci are also problematic for bioMérieux VITEK MALDI-TOF MS. Rychert and colleagues were unable to differentiate between Streptococcus pyogenes (Group A Streptococcus , GAS) and Streptococcus dysgalactiae with only 51% of S. dysgalactiae correctly identified.

Gram-positive rods

Prior to the implementation of MALDI-TOF MS, clinical microbiology laboratories did not have a reliable and efficient method to identify most gram-positive rods ( Fig. 84.7 ). This was particularly problematic for the Corynebacterium spp. and to a lesser extent, the Bacillus spp. These organisms are increasingly recognized as causes of opportunistic infections in some patient populations, including immunocompromised patients.

Bacillus spp. are a group of closely related organisms and there is little evidence evaluating the ability of MALDI-TOF MS to produce accurate identifications of these organisms. Of significance for any laboratory is the ability to recognize B. anthracis, which is a member of the B. cereus group. Most commonly used MALDI-TOF MS systems do not include B. anthracis in commercially available, US Food and Drug Administration (FDA)-cleared databases and therefore are unable to correctly identify this organism. For specialized databases that do include B. anthracis, MALDI-TOF MS can accurately identify this organism. It should be noted that routine identification of B. anthracis using MALDI-TOF MS is not recommended. Instead, Rule Out/Refer procedures from the Centers for Disease Control and Prevention (CDC) and the Association of Public Health Laboratories (APHL) should be followed.

Less information is available regarding the ability of MALDI-TOF MS to identify other gram-positive rods. Studies of isolates including Arcanobacterium haemolyticum, Trueperella bernardiae, Brevibacterium casei, Microbacterium spp., Arthrobacter spp., Rothia spp., Dermabacter hominis, Rhodococcus equi, and Listeria monocytogenes found that most of these isolates were correctly identified to the species level and none were incorrectly identified. ^,

Gram-negative bacilli and coccobacilli

The ability of MALDI-TOF MS to accurately identify gram-negative bacilli and coccobacilli is well documented in the peer-reviewed literature. MALDI-TOF MS performance for Enterobacterales , glucose nonfermenting organisms, and fastidious gram-negative bacilli will be discussed.

Enterobacterales

Organisms within the Enterobacterales order are some of the most commonly encountered pathogens in the clinical microbiology laboratory ( Fig. 84.8 ). Traditional biochemical identification methods work relatively well for this group of organisms, and as a result, many laboratories may continue to use these methods in lieu of MALDI-TOF MS. In particular, many laboratories use “limited identification” strategies for Escherichia coli , as outlined in the CLSI document for the abbreviated identification of bacteria and yeast. According to this document, E. coli can be identified with indole (positive), oxidase (negative), lactose fermentation, and ß-hemolytic growth. Notably, approximately 5% of E. coli do not ferment lactose or are nonhemolytic. For nonhemolytic colonies, a negative pyrrolidonyl arylamidase (PYR) test can confirm the presence of E. coli . Nonhemolytic and lactose nonfermenters can be confirmed as E. coli with a positive rapid methylumbelliferyl-ß- d -glucuronidase (MUG) test.

E. coli identification deserves special mention in this section because the inability of MALDI-TOF MS to distinguish E. coli from Shigella is probably its greatest weakness. In most systems, Shigella is not present in the database because its spectra are indistinguishable from those of E. coli. As a result, the default identification from MALDI-TOF MS is E. coli for either organism. It is particularly important to accurately identify Shigella spp. because there are significant public health implications associated with Shigella spp. infection. In addition, MALDI-TOF MS cannot provide Shigella spp. serotype information, thus laboratories may wish to retain the ability to serotype using conventional methods. Importantly, MALDI-TOF MS can reliably differentiate E. coli from non- Shigella spp. Enterobacterales . As a result, the method can be used confidently for the identification of isolates from specimens where the isolation of Shigella spp. would be a rare event. For other common clinical isolates of Enterobacterales , MALDI-TOF MS systems perform reliably, with misidentification rates of less than 1%. ^, Of note, MALDI-TOF MS identification of Yersinia pestis may pose a challenge, but the data are mixed in this regard. ^,

Glucose nonfermenting organisms

One of the greatest benefits of MALDI-TOF MS is its ability to reliably identify glucose nonfermenting organisms ( Fig. 84.9 ). MALDI-TOF MS accurately identifies Pseudomonas aeruginosa , arguably the most clinically important glucose nonfermenting organism. Identification of isolates cultured from respiratory specimens of CF patients has improved markedly with the use of MALDI-TOF MS. Because MALDI-TOF MS identification is less affected by the growth conditions of an organism than traditional methods, as mentioned previously, it can reliably identify organisms with altered growth characteristics, such as those in CF patients. B. cepacia complex organisms are significant pathogens in CF patients, and accurate identification is vital to their proper management. An analysis of the accuracy of MALDI-TOF MS identification within the B. cepacia found that all systems were able to correctly identify members within the B. cepacia complex. ^, Importantly, MALDI-TOF MS can accurately identify other glucose nonfermenting, colistin-resistant organisms ( Pandorea and Ralstonia ), which have historically been difficult to confidently differentiate from B. cepacia complex. Of note, MALDI-TOF MS systems can have difficulty identifying Burkholderia mallei and B. pseudomallei. These are agents of bioterrorism that are not included in most FDA-cleared MALDI-TOF MS databases.

Fastidious gram-negative bacilli

This general category of organisms includes significant diversity and a detailed discussion of MALDI-TOF MS performance for every fastidious gram-negative rod is outside the scope of this chapter. Rather, this discussion focuses on the more commonly encountered species ( Fig. 84.10 ).

MALDI-TOF MS performance for the identification of fastidious gram-negative rods has excellent accuracy. Of the HACEK group organisms, a group of species that can cause subacute endocarditis (will be discussed in more detail later in the chapter), Haemophilus parainfluenzae, Aggregatibacter actinomycetemcomitans, Aggregatibacter aphrophilus (formerly Haemophilus aphrophilus ) , Aggregatibacter paraphrophilus (formerly Haemophilus paraphrophilus ) , and Eikenella corrodens were accurately identified to the species level. Limited data from other studies suggest that MALDI-TOF MS is able to accurately identify Kingella kingae, as well as Cardiobacterium hominis . ^,

Practically speaking, genus-level identification is probably sufficient for many fastidious gram-negative rods. For example, it is probably not necessary to differentiate Dysgonomonas capnocytophagoides from D. mossii. Conversely, it could be clinically important to differentiate species within the Haemophilus genera or the Bordetella genera. Fortunately, MALDI-TOF MS appears readily able to identify most of these genera to the species level.

Curved gram-negative rods

MALDI-TOF MS does well identifying many species of Campylobacter , the most common curved gram-negative rods encountered in the clinical microbiology laboratory. Multiple studies have shown the ability of MALDI-TOF MS to accurately differentiate commonly encountered Campylobacter spp., including C. jejuni , and C. coli . ^, Others have shown that MALDI-TOF MS can discriminate between Vibrio cholerae and aeromonads, which can sometimes mimic V. cholerae in stool culture.

Gram-negative diplococci

There is comparatively little peer-reviewed data evaluating the performance of the MALDI-TOF MS and its ability to accurately identify gram-negative diplococci ( Neisseria spp. and Moraxella catarrhalis ) (see Fig. 84.10 ). M. catarrhalis is readily identified with MALDI-TOF MS, but some studies have found MALDI-TOF MS struggles with differentiation of pathogenic Neisseria meningitidis from nonpathogenic Neisseria spp., such as Neisseria polysaccharea. ^, In addition, MALDI-TOF MS may not accurately differentiate among nonpathogenic Neisseria spp. such as Neisseria subflava, Neisseria sicca, and Neisseria cinerea.

In most circumstances, it is not clinically relevant to differentiate among nonpathogenic Neisseria spp. However, it is extremely important that laboratories be able to accurately identify N. gonorrhoeae and N. meningitidis , not only so that laboratories can properly guide patient care, but also, in the case of N. meningitidis, so that laboratorians can protect themselves from exposure to a potentially life-threatening infection. Although laboratory-acquired N. meningitidis infection is rare, it results in high morbidity and approximately 50% mortality. Given the inability of MALDI-TOF MS to reliably identify some Neisseria spp. isolates, it is critical that laboratories maintain back-up methods for confirmatory identification. These back-up methods include 16S rDNA sequencing (not widely available), biochemical reactions, such as gamma-glutamyltransferase activity, and production of acid from sucrose, or from glucose and maltose on cysteine-rich media.

Anaerobes

Anaerobic infections are frequently polymicrobic, as they result from either endogenous dissemination of commensal microbiota or from traumatic penetration, which introduces environmental anaerobes to ordinarily sterile body sites. Before implementation of MALDI-TOF MS, laboratory methods failed to reliably identify many anaerobic organisms. However, MALDI-TOF MS can rapidly and accurately identify many important anaerobic organisms including Clostridium spp., Bacteroides spp., Cutibacterium (formerly Propionibacterium ) spp., Peptostreptococcus spp. and related organisms, Fusobacterium spp., Actinomyces spp., and Lactobacillus spp. ^, Similar to other bacterial categories, MALDI-TOF MS may not identify all anaerobes, but it rarely generates erroneous results.

MALDI-TOF MS has become so effective in identifying anaerobic organisms that many laboratories have altered their traditional protocols such that routine aero-tolerance testing (comparative growth studies to determine whether an organism is an obligate anaerobe) is no longer routinely performed. Before MALDI-TOF MS, laboratories would subculture suspected anaerobes to both aerobic and anaerobic conditions and compare the rates of growth. Strict anaerobes would be confirmed by growing under anaerobic but not aerobic conditions. This procedure delays results because it requires at a minimum overnight incubation. Despite the ease and accuracy of MALDI-TOF MS identification of anaerobes, it is important that laboratories maintain a working knowledge of some basic organism characteristics such that MALDI-TOF MS results can be questioned and verified. Important identifying characteristics include the Gram stain, morphology, colony pigment, antibiotic resistance profiles, and fluorescence. See Fig. 84.11 for an outline of clinically important anaerobes and their defining characteristics.

Aerobic actinomycetes

Aerobic actinomycetes are made up of a broad group of gram-positive bacillary organisms that are often characterized by the presence of rudimentary branching. A positive modified acid-fast stain is also considered a key indicator among many genera of this group. Although over 40 genera have been classified as part of the aerobic actinomycete group, only a small minority are associated with human infections— Nocardia , Rhodococcus , Gordonia , Tsukamurella , Actinomadura , Streptomyces , and Tropheryma . Other members of this group such as Amycolata, Amycolatopsis, Pseudonocardia, Dermatophilus, Dietzia, and Nocardiopsis , although occasionally isolated from patient specimens, are saprophytes and should almost always be considered contaminants or colonizers.

Aerobic actinomycetes take up to two weeks to culture. Hence, infections due to these organisms are often overlooked as they are frequently overgrown by other microbiota associated with the site of infection and/or they require extended incubation time to grow the organism. Historically the identification of aerobic actinomycetes required evaluating both microscopic and macroscopic morphologies, in addition to testing for the metabolism of glucose, arylsulfatase production, growth in lysozyme, and determining the absence or presence of mycolic acid, a cell wall component. These methods are laborious and time consuming and are only able provide a presumptive or genus-level identification in most cases. ^,

The use of MALDI-TOF MS for identification of actinomycetes has both shortened the time to identification and allowed for species-level identification that would have necessitated the need for DNA sequencing in the past. ^, However, unlike most bacteria identified using MALDI-TOF MS, aerobic actinomycetes may require additional steps prior to spotting to the target slide. These inactivation steps entail the use of glass beads with ethanol to disrupt the cell wall, and formic acid and acetonitrile to complete the extraction of the proteins. Both the MALDI Biotyper (Bruker) and the VITEK MS (bioMérieux) can identify most pathogenic aerobic actinomycetes (i.e., Nocardia , Rhodococcus , Gordonia , Tsukamurella , Actinomadura , and Streptomyces ), with the only limitation being the robustness of the MALDI-TOF MS database. ^{, ,}

Sequencing basics

Overview of chain-termination DNA sequencing

This sequencing method, used by many laboratories, is a variation of the Sanger chain-termination technique developed two decades ago. The method is reviewed in detail in Chapter 64 . In brief, DNA extracted from the bacterial isolate serves as a template to which specific, short, complementary oligonucleotides (i.e., universal primers complementary to conserved regions of the 16S rDNA) anneal and initiate DNA synthesis. Newly created fragments are complementary to the template DNA. In the following cycle-sequencing reaction, the purified PCR products serve as the template, and forward or reverse primers are added to a mixture containing deoxyribonucleoside triphosphates (dNTPs) and four different, fluorescently labeled, dideoxyribonucleoside triphosphates (ddNTPs; dye terminators), Taq polymerase, and buffer. During this cycling reaction, DNA is denatured, the forward or reverse primer anneals and is extended such that sufficient fluorescent-labeled DNA is generated for subsequent sequencing, and during the extension step, chain termination occurs. In the reaction mix, dNTPs and ddNTPs are at concentrations such that some fluorescently labeled ddNTP will be incorporated instead of a dNTP at each nucleoside position in newly generated fragments. This results in terminated products of varying lengths. During capillary electrophoresis, these fragments separate by length, and because the terminating nucleoside is known (fluorescently labeled ddNTP), the sequencing machine will detect changes in fluorescence and use these changes to generate the nucleic acid sequence. The two strands of DNA are sequenced separately, generating forward and reverse complementary sequences thereby allowing for resolution of any base ambiguities or other errors. A generated DNA sequence (i.e., consensus sequence) is assembled by aligning the forward and reverse sequences, and then compared to a database using analysis software. Following query of the database, identification of the unknown bacteria can be determined. See At a Glance feature for an overview of sequencing. The utility of 16S rDNA sequencing as a tool in microbial identification depends on two key elements: correctly identifying or “labeling” each sequence and deposition of complete, unambiguous nucleotide sequences into public or private databases. There are a number of well-known databases of 16S rDNA sequences, many of which contain high-quality, current, and optimally curated sequence data. For example, GenBank, which contains the largest database of nucleotide sequences, contains over 90,000 deposited 16S rDNA sequences.

AT A GLANCE

Typically, nucleotide sequences are reported in terms of percent identity . This phrase refers to the number of identical nucleotide bases shared by the query and reference sequences divided by the number of nucleotide bases sequenced. Interpretive criteria were published by the Clinical and Laboratory Standards Institute (CLSI) in 2008 and updated in 2018. An additional CLSI document addresses diagnostic sequencing using capillary-based sequencers and other topics such as isolation and extraction of nucleic acid; template preparation; sequence generation, alignment, and assembly; validation and verification; ongoing quality assurance; and reporting of results.

Finally, there are primer candidates for broad-range amplification involving the V1 to V3 area of the 16S rDNA that do not have significant cross-reactivity with human DNA. By using these primers and other advances, such as dual priming oligonucleotides, the detection of bacterial 16S rDNA sequences directly from human clinical material is now possible. ^{, ,}

Potential pitfalls of sequence-based identification

Although it is clear that 16S rDNA sequencing has a significant role in the identification of bacteria in the clinical laboratory, it is not foolproof or applicable in every situation. Misidentification of bacteria by 16S rDNA sequencing has been reported for organisms such as S. pneumoniae, E. coli, and B. anthracis . Therefore it is important to correlate sequence-based findings with the phenotypic traits of the organism being characterized. In addition, various software packages can give different results. Finally, because unusual taxa are detected and reported by gene sequencing, it is important for the laboratory to provide additional information to help clinicians in placing the new species in a familiar context.

PCR and non-PCR nucleic acid amplification tests

Historically the use of nucleic acid amplification tests (NAATs) has been associated with basic science research rather than the clinical laboratory setting, as the development of these “high-complexity” tests required a skilled work force that understood molecular methods and a laboratory space that could assure that the multi-step process of testing would limit contamination. However, advances in molecular methods, such as transcription mediated amplification (TMA), real-time polymerase chain reaction (RT-PCR), and quantitative PCR (qPCR), has allowed for the commercialized development of sample-to-result instruments, thereby eliminating the need for highly skilled technologists and a dedicated space for some molecular testing.

Most of the nucleic acid-based tests used in the detection of bacteria are qualitative assays. Typically, these qualitative tests are used to identify pathogens from a sterile source, such as blood or CSF. On nonsterile sources, these qualitative tests focus on definitive pathogens, such as Chlamydia trachomatis or Neisseria gonorrhoeae , which are not part of normal microbiota and where a positive result would be clinically actionable. The use of NAATs has also proven useful in identifying organisms that require extended incubation times (e.g., Mycoplasma pneumoniae ), selective media (e.g., Bordetella pertussis/parapertussis ), or cell lines to grow (e.g., C. trachomatis ). The use of molecular assays has also had practical effects in the appropriate management of antimicrobial therapy. Examples of patient management being influenced by molecular testing include patients being screened for methicillin-resistant Staphylococcus aureus (MRSA) colonization to determine if empiric vancomycin therapy should be discontinued in patients presenting with a lower respiratory tract infection or if decolonization measures should be instituted prior to surgery. Furthermore, these molecular-based tests have practical implications from an infection prevention perspective as well, as patients colonized or infected with MRSA, vancomycin-resistant Enterococcus (VRE), or carbapenem-resistant Enterobacterales (CRE) can be rapidly placed on appropriate precaution measures, thus limiting the subsequent transmission of these organisms.

Although NAATs have many positive aspects, one of the key constraints to molecular-based testing, compared to culture, is its inability to detect the organism that is causing the infection unless the correlative primer set is present in the assay. Alternatively, from a clinician’s perspective it can be just as difficult to determine the etiologic agent based simply on the signs and symptoms of the patient, which leads to the ordering of multiple culture types (i.e., aerobic/anaerobic bacterial cultures, fungal cultures, and/or AFB cultures) to assure that the pathogen of interest can be identified. This dilemma has been partially resolved with the advent of multiplex syndromic panels, where the selection of targets is based on pathogens most commonly associated with a specific infectious disease syndrome, such a bacteremia, meningitis/encephalitis, or gastroenteritis. The implementation of these syndromic panels has allowed for faster turnaround times, greater sensitivity, and significantly better patient outcomes as patients can now be actively managed in real time. For additional details on molecular-based testing and its role in the clinical microbiology setting, refer to Chapter 67 .

Direct specimen testing for pathogen identification

In recent years, direct specimen testing for pathogen identification has become more commonplace in clinical laboratories. These technologies include, but are not limited to, broad-range 16S rDNA PCR and sequencing, magnetic resonance testing, and other near-direct specimen testing methods, such as direct MALDI-TOF MS identification of positive blood cultures, and rapid culture-based systems for pathogen identification and susceptibility testing from positive blood cultures.

Broad-range sequencing is a method of pathogen identification whereby a patient specimen is submitted for 16S rDNA PCR (or other targets, such as hsp65 and rpoB for mycobacteria) and sequencing of the PCR product. There are a number of situations where this technology could add tremendous clinical value. However, this testing should be used with discretion, as it is often costly and, in some instances, unnecessary.

One of the advantages of broad range 16S rDNA sequencing is the ability to identify bacteria present in normally sterile specimens, like tissue and CSF. These identifications are unbiased with respect to viability and cultivability of the organism. Indeed, this aspect of this testing is useful, especially when specimens are inadvertently placed into formalin prior to microbiological culture, as formalin-fixed paraffin-embedded tissue is generally acceptable for this testing, especially if potential pathogens are noted in the pathology report. Challenges associated with broad-range sequencing include contamination, long turnaround times, and the inability to obtain antimicrobial susceptibility testing. An in-depth review of broad-range sequencing is beyond the scope of this chapter. ^,

Magnetic resonance technology has come into use in some clinical microbiology laboratories in the past decade. T2 Biosystems (Lexington, MA) currently has two FDA-cleared panels on their T2Dx system for direct detection of pathogens from whole blood specimens, for both candidemia and bacteremia. This technology leverages the ability of magnetic resonance to track changes in water molecule interactions. Recent studies of this technology show very good test performance characteristics for the targets on the test panel (90% for both sensitivity and specificity), but low overall performance for all bacteremia (43%). ^, However, many questions about the utility of this test remain, particularly in regard to how the results are used for patient management and antimicrobial stewardship.

In addition to the direct specimen testing assays, described above, there are a number of near-direct methods available for identification and susceptibility testing from positive blood culture broth, including rapid culture-based and molecular FDA-cleared assays, in addition to laboratory-developed tests that identify and test bacteria purified from positive blood culture broth by MALDI-TOF MS and/or various antimicrobial susceptibility systems. These methods, especially those that provide direct susceptibility testing from blood culture broth, can provide clinically actionable information in a much shorter time frame than typical culture and susceptibility testing.

Preanalytical considerations

Analytical considerations