Chapter S13: Laboratory Diagnosis of Infectious Diseases

The laboratory diagnosis of infection requires the demonstration—either direct or indirect—of viral, bacterial, fungal, or parasitic agents in tissues, fluids, or excreta of the host. Clinical microbiology laboratories are responsible for processing these specimens and also for determining the antibiotic susceptibility of bacterial and fungal pathogens. Traditionally, detection of pathogenic agents has relied largely on either the microscopic visualization of pathogens in clinical material or the growth of microorganisms in the laboratory. Identification has been—and often still is—based on phenotypic characteristics such as fermentation profiles for bacteria, cytopathic effects in tissue culture for viral agents, and microscopic morphology for fungi and parasites. These techniques are reliable but are often time-consuming. Increasingly, protein analysis or genotypic techniques are becoming standard methods for detection, quantitation, and/or identification of microorganisms in the clinical microbiology laboratory, replacing phenotypic characterization and microscopic visualization methods. This chapter discusses general concepts of diagnostic testing, with an emphasis on detection of bacteria. Detection of viral, fungal, and parasitic pathogens is discussed in greater detail in separate chapters (see Chaps. 185, 206, and 217, respectively).

Microorganisms can be detected by automated or manual methods that depend on the presence of molecules unique to the specific pathogen. Molecules useful for detection of microorganisms include structural components of bacteria, fungi, and viruses; specific antigens; metabolic end products; unique DNA or RNA sequences; enzymes; toxins or other proteins; and surface polysaccharides. These are detected by a variety of methods, including immunofluorescence; chemiluminescence for DNA/RNA probes; flame ionization detection of short- or long-chain fatty acids; and detection of substrate use or end-product formation as color changes, of enzyme activity as a change in light absorbance, of turbidity changes as a measure of growth, of cytopathic effects in cell lines, and of particle agglutination as a measure of antigen presence. Amplification enhances the sensitivity with which weak signals can be detected. Growth of a single bacterium or yeast into a visible colony on an agar plate is a common method of amplification. More rapid amplification can be achieved with techniques such as polymerase chain reaction (PCR), ligase chain reaction (LCR), and transcription-mediated amplification (TMA), all of which target the pathogen’s DNA/RNA; enzyme immunoassays (EIAs) for antigens; electronic amplification (for gas–liquid chromatography assays); antibody capture methods (for concentration and/or separation); and selective filtration or centrifugation.

Direct detection refers to detection of pathogens without the use of culture. Molecular methods of direct detection are discussed below.

MICROSCOPY AND STAINING

The field of microbiology was defined largely by the development and use of the microscope. The examination of specimens by microscopic methods rapidly provides useful diagnostic information. Staining techniques permit organisms to be seen more clearly.

The simplest method for microscopic evaluation is the wet mount, which is used, for example, to examine cerebrospinal fluid (CSF) for the presence of Cryptococcus neoformans, with India ink as a background against which to visualize large-capsuled yeast cells. Wet mounts with dark-field illumination also are used to detect spirochetes in genital lesions and Borrelia or Leptospira in blood. Skin scrapings and hair samples can be examined with the use of either 10% KOH wet-mount preparations or the calcofluor white method and ultraviolet illumination to detect fungal elements as fluorescing structures. Staining of wet mounts—e.g., with lactophenol cotton blue for fungal elements—often is used for morphologic identification.

Bacteria are difficult to see by light microscopy unless they are stained. Although simple one-step stains can be used, differential stains are more common.

Gram’s Stain

Gram’s stain differentiates organisms with thick peptidoglycan cell walls (gram-positive) from those with thin peptidoglycan cell walls and outer membranes that can be dissolved with alcohol or acetone (gram-negative). Morphology and Gram’s stain characteristics often can be used to categorize stained organisms into groups such as streptococci, staphylococci, and clostridia (Table S13-1). Gram’s stain is particularly useful for examining sputum for polymorphonuclear leukocytes (PMNs) and bacteria. Sputum specimens from immunocompetent patients with ≥25 PMNs and <10 epithelial cells per low-power field often provide clinically useful information. However, the presence in “sputum” samples of >10 epithelial cells per low-power field and of multiple bacterial types suggests contamination with the oral microflora. Despite the difficulty of discriminating between normal microflora and pathogens, Gram’s stain may prove useful for specimens from areas with a large resident microflora. Gram’s staining of vaginal swab specimens can be used to detect epithelial cells covered with gram-positive bacteria in the absence of lactobacilli and the presence of gram-negative rods—a result that suggests bacterial vaginosis. Detection of gram-negative cocci in pairs in meatal pus from a man with dysuria is diagnostic of gonorrhea.

The examination of samples from normally sterile body sites (e.g., CSF or joint, pleural, or peritoneal fluid) with Gram’s stain is useful for determining whether bacteria and/or PMNs are present. The sensitivity is such that >10⁴ bacteria/mL should be detected. Centrifugation often is performed before staining to concentrate specimens thought to contain low numbers of organisms. This simple method is particularly useful for examination of CSF for bacteria and white blood cells or examination of sputum for mycobacteria.

Acid-Fast Stain

The acid-fast stain identifies acid-fast bacteria (AFB; mycobacteria) by their retention of carbol fuchsin dye after acid/organic solvent disruption. Modifications of this procedure allow the differentiation of Actinomyces from Nocardia or other weakly (or partially) acid-fast organisms. The acid-fast stain is applied to sputum, other fluids, and tissue samples when Mycobacterium species are suspected. Because few AFB may be present in an entire smear, even when the specimen has been concentrated by centrifugation, identification of the pink/red AFB against the blue background of the counterstain requires a trained eye. An alternative method for detection of AFB is the auramine–rhodamine fluorescent dye stain.

Fluorochrome Stains

Fluorochrome stains such as acridine orange are used to identify white blood cells, yeasts, and bacteria in body fluids. Capsular, flagellar, and spore stains are used for identification or demonstration of characteristic structures.

Immunofluorescent Stains

The direct immunofluorescent antibody technique uses antibody coupled to a fluorescent compound (e.g., fluorescein) and directed at a specific antigenic target to visualize organisms. When samples are examined under appropriate conditions, the fluorescing compound absorbs ultraviolet light and re-emits light at a higher wavelength that is visible to the human eye. The indirect immunofluorescent antibody technique uses an unlabeled (target) antibody to bind a specific antigen. The specimen is then stained with a fluorochrome-labeled antibody directed at the target antibody. Because each unlabeled target antibody attached to the appropriate antigen has multiple sites for attachment of the second antibody, the visual signal is amplified. Immunofluorescence is used to detect viral antigens (e.g., cytomegalovirus, herpes simplex virus, and respiratory viruses) within cultured cells or clinical specimens as well as many difficult-to-grow bacterial agents (e.g., Legionella pneumophila) in clinical specimens.

Latex agglutination assays and EIAs are rapid and inexpensive methods for identifying organisms, extracellular toxins, and viral agents by means of protein and polysaccharide antigens. Such assays may be performed directly on clinical samples or after growth of organisms on agar plates or in liquid media. Antibodies coupled to a reporter (such as latex particles or an enzyme) are used for detection of antibody–antigen binding reactions.

Direct agglutination of bacterial cells with specific antibody is simple but relatively insensitive; latex agglutination and EIAs are more sensitive. Some cell-associated antigens, such as capsular polysaccharides and lipopolysaccharides, can be detected by agglutination of a suspension of bacterial cells when antibody is added; this method is useful for typing of the somatic antigens of Shigella and Salmonella. EIAs employ antibodies coupled to an enzyme, and an antigen–antibody reaction results in the conversion of a colorless substrate to a colored product. Most of these assays provide information about whether antigen is present but do not quantify the antigen. EIAs are also useful for detecting bacterial toxins—e.g., toxins produced by Shiga toxin–producing Escherichia coli.

Rapid and simple immunoassays for antigens of group A Streptococcus, influenza virus, and respiratory syncytial virus can be used in the clinical setting without a specialized diagnostic laboratory. Such tests usually are reasonably specific but have only modest sensitivity.

Measurement of serum antibody provides an indirect marker for past or current infection with a specific viral agent or other pathogens, including Brucella, Legionella, Rickettsia, and Helicobacter pylori. Serologic methods can be used to determine whether an individual has protective antibody levels or is infected by a specific pathogen. Determination of an antibody level as a measure of current immunity is important in the case of viral agents for which there are vaccines, such as rubella virus and varicella-zoster virus; assays for this purpose normally use one or two dilutions of serum for a qualitative determination of protective antibody levels. Quantitative serologic assays to detect increases in antibody titers most often employ paired serum samples obtained at the onset of illness and 10–14 days later (i.e., acute- and convalescent-phase samples). Since the incubation period before symptoms are noted may be long enough for an antibody response to occur, the demonstration of acute-phase antibody alone is often insufficient to establish the diagnosis of active infection as opposed to past exposure. A fourfold increase in total antibody titer between the acute- and convalescent-phase samples is regarded as evidence for active infection. In addition, IgM may be useful as a measure of an early, acute-phase antibody response. For certain viral agents, such as Epstein-Barr virus, the antibodies produced may be directed at different antigens during different phases of the infection. For this reason, most laboratories test for antibody directed at both viral capsid antigens and antigens associated with recently infected host cells to determine the stage of infection.

SPECIMEN COLLECTION AND TRANSPORT

To culture bacterial, fungal, or viral pathogens, an appropriate sample must be placed into the proper medium for growth. The success of efforts to identify a specific pathogen often depends on the collection and transport process coupled to a laboratory-processing algorithm suitable for the specific sample and suspected pathogen(s). In some instances (e.g., when urethral swabs are being cultured for Neisseria gonorrhoeae), it is better for specimens to be plated at the time of collection rather than first being transported to the laboratory. In general, the more rapidly a specimen is plated onto appropriate media, the better the chance is for isolating bacterial pathogens. Deep tissue or fluid (pus) samples are more likely to give useful culture results than are superficial swab specimens. Table S13-2 lists procedures for collection and transport of common specimens. Because there are many pathogen-specific paradigms for these procedures, it is important to seek advice from the microbiology laboratory when in doubt about a particular situation.

ISOLATION OF BACTERIAL PATHOGENS

Isolation of pathogens from clinical material relies on the use of culture media that support bacterial growth. Such media are composed of agar, nutrients, and sometimes substances that inhibit the growth of other bacteria. Broth is employed for growth of organisms from specimens with few bacteria, such as peritoneal dialysis fluid or CSF, or from samples in which anaerobes or other fastidious organisms may be present. Broths that allow the growth of small numbers of organisms may be subcultured onto solid medium once growth is detected. The use of liquid medium for all specimens is not worthwhile.

Two basic strategies are used to isolate pathogenic bacteria. The first is to employ enriched media that support the growth of any bacteria that may be present at a site that is normally sterile, such as blood or CSF. The second strategy is to use selective media to isolate specific pathogenic bacterial species from samples that contain a complex population of bacteria (e.g., stool or genital tract secretions). Antimicrobial agents or other substances are incorporated into the agar medium to inhibit the growth of all but the bacteria of interest. After incubation, organisms that grow on such medium are characterized further to determine whether they are pathogens (Fig. S13-1).

BLOOD CULTURE

The detection of microbial pathogens in blood is difficult because the number of organisms present in the sample is often low and the organisms’ ability to replicate may be reduced by humoral defense mechanisms or antimicrobial agents. Automated blood culture systems detect production of gas (mainly CO₂) by bacteria or yeasts growing in broth in a blood culture bottle. Because the bottles are monitored frequently, a positive culture often is detected more rapidly than by manual techniques, and important information, including the results of Gram’s stain and preliminary susceptibility assays, can be obtained sooner. Several factors affect the yield of blood culture from bacteremic patients. Increasing the volume of blood tested increases the chance of a positive culture. For example, a sample-size increase from 10 to 20 mL of blood increases the proportion of positive cultures by ~30%, although this effect is less pronounced in patients with bacterial endocarditis. Obtaining multiple samples for culture (up to three per 24-h period) also increases the chance of detecting a bacterial pathogen. Prolonged culture and blind subculture for detection of most fastidious bacteria (e.g., HACEK organisms) are not needed with automated blood culture systems.

Automated systems also have been applied to the detection of microbial growth from specimens other than blood, such as peritoneal and other normally sterile fluids. Mycobacterium species can be detected in certain automated systems if appropriate liquid media are used for culture. Although automated blood culture systems are more sensitive than lysis–centrifugation methods (e.g., Isolator) for yeasts and most bacteria, lysis–centrifugation culture is recommended for filamentous fungi, Histoplasma capsulatum, and some fastidious bacteria (Legionella and Bartonella).

Once bacteria are isolated, characteristics that are readily detectable after growth on agar media (colony size, color, hemolytic reactions, odor, microscopic appearance) may suggest a species, but definitive identification requires additional tests. Identification methods include classic biochemical phenotyping, which is still the most common approach, as well as more sophisticated methods such as mass spectrometry, gas chromatography, and nucleic acid tests (see below).

BIOCHEMICAL PHENOTYPING

Classic biochemical identification of bacteria entails tests for protein or carbohydrate antigens, the production of specific enzymes, the ability to metabolize specific substrates and carbon sources (such as carbohydrates), or the production of certain metabolites. Rapid versions of some of these tests are available, and many common organisms can be identified on the first day of growth. Other organisms, particularly gram-negative bacteria, require more extensive testing by either manual or automated methods.

Automated systems allow rapid phenotypic identification of bacterial pathogens. Most of these systems are based on biotyping techniques in which isolates are grown on multiple substrates and the reaction pattern is compared with known patterns for various bacterial species. This procedure is relatively fast. Commercially available systems include miniaturized fermentation, coding to simplify recording of results, and probability calculations for the most likely pathogens. If the biotyping approach is automated and the reading process is coupled to computer-based analysis of data, rapidly growing organisms (such as Enterobacteriaceae) can be identified within hours of detection on agar plates.

Several systems use substrates to detect bacterial enzymes for organism identification within 2–3 h. Rather than relying on bacterial growth, these systems detect enzymes present in a heavy inoculum of bacteria. In addition, some systems use fluorogenic substrate/end-product detection methods to increase sensitivity through signal amplification.

GAS–LIQUID CHROMATOGRAPHY

Gas–liquid chromatography is used to detect metabolic end products of bacterial fermentation. One common application is identification of short-chain fatty acids produced by obligate anaerobes during glucose fermentation. Because the types and relative concentrations of volatile acids differ among the various genera and species that make up this group of organisms, such information serves as a metabolic fingerprint for a particular isolate. Gas–liquid chromatography can be coupled to a sophisticated signal-analysis software system for identification and quantitation of long-chain fatty acids in the outer membranes and cell walls of bacteria and fungi. For any particular species, the types and relative concentrations of long-chain fatty acids are distinctive enough to allow differentiation even from closely related species.

MATRIX-ASSISTED LASER DESORPTION/IONIZATION TIME-OF-FLIGHT MASS SPECTROMETRY (MALDI-TOF MS)

MALDI-TOF MS is a rapid and accurate method for identifying microorganisms by protein analysis. The organism is mixed with a chemical matrix, dried on a target plate, and then pulsed with a laser. The laser ionizes and vaporizes the microbial proteins, which then travel through a charged vacuum chamber to a detector. The time of flight to the detector is measured for the individual proteins, and the resulting pattern (or fingerprint) is compared with a library of known patterns for various microorganisms to identify the test organism. The primary molecules detected using MALDI-TOF MS are the ribosomal proteins, which are the most abundant proteins in the cell.

The primary clinical advantage of MALDI-TOF MS is that it takes only minutes to identify an organism directly from a colony on an agar plate. Conventional methods of identification take an additional 1–3 days; yeasts and anaerobes take longest. MALDI-TOF MS is very accurate for identification of common bacteria and yeasts. However, some less common organisms are absent from the U.S. Food and Drug Administration (FDA)–approved databases used by MALDI-TOF MS devices, and these organisms cannot be identified unless the laboratory has validated use of another database. MALDI-TOF MS is used for identification of hyphal molds in a small number of laboratories but is not widely available. It should be noted that MALDI-TOF cannot currently differentiate some closely related organisms, such as Shigella and E. coli. Thus other methods must be employed to confirm these identifications. Other potential uses of MALDI-TOF MS include identification of bacteria and yeasts directly from low-protein clinical specimens (e.g., urine), detection of β-lactamase activity, and strain typing of bacteria; however, these applications are still under development.

NUCLEIC ACID TESTS

Techniques for the detection and quantitation of specific DNA and RNA base sequences in clinical specimens have become powerful tools for the diagnosis of bacterial, viral, parasitic, and fungal infections. Nucleic acid tests are used for four purposes. First, they are used to detect, and sometimes to quantify, specific pathogens in clinical specimens. Second, such tests are used for identification of organisms (usually bacteria) that are difficult to identify by conventional methods. Third, nucleic acid tests are used to determine whether two or more isolates of the same pathogen are closely related (i.e., whether they belong to the same clone or strain). Fourth, these tests are used to predict the sensitivity of organisms to a small number of chemotherapeutic agents. Current technology encompasses a wide array of methods for amplification and signal detection, some of which have been approved by the FDA for clinical diagnosis. Many laboratories have developed their own nucleic acid tests for pathogens. These must be rigorously validated before they can be used for patient care.

Use of nucleic acid tests generally involves lysis of intact cells or viruses and denaturation of the DNA or RNA to render it single-stranded. The probe(s) or primer(s) complementary to the pathogen-specific target sequence are hybridized to the target sequence in a solution or on a solid support, depending on the system employed. In situ hybridization of a probe to a target also is possible and allows the use of probes with agents present in tissue specimens. Once the probe(s) or primer(s) have been hybridized to the target (biologic signal), a variety of strategies may be employed to detect, amplify, and/or quantify the target–probe complex (Fig. S13-2).

Probes for Direct Detection of Pathogens in Clinical Specimens

Nucleic acid probes are used for direct detection of pathogens in clinical specimens without amplification of the target strand of DNA or RNA. Such tests detect a relatively short sequence of bases specific for a particular pathogen on single-stranded DNA or RNA by hybridization of a complementary sequence of bases (probe) coupled to a reporter system that serves as the signal for detection. Nucleic acid probes are commercially available for direct detection of various bacterial and parasitic pathogens, including Chlamydia trachomatis, N. gonorrhoeae, and group A Streptococcus. A combined assay to detect and differentiate agents of vaginitis/vaginosis (Gardnerella vaginalis, Trichomonas vaginalis, and Candida species) also has been approved. An assortment of probes is available for confirming the identity of cultured pathogens, including some dimorphic molds, Mycobacterium species, and other bacteria (e.g., Campylobacter species, Streptococcus species, and Staphylococcus aureus). Probes for the direct detection of bacterial pathogens often are aimed at highly conserved 16S ribosomal RNA sequences, of which there are many more copies than there are of any single genomic DNA sequence in a bacterial cell. The sensitivity and specificity of probe assays for direct detection are comparable to those of more traditional assays, including EIA and culture, but probe assays are less sensitive than nucleic acid amplification tests (NAATs, see below).

In an alternative probe assay called hybrid capture, an RNA probe anneals to a DNA target, and the resulting DNA/RNA hybrid is captured on a solid support by antibody specific for DNA/RNA hybrids (concentration/amplification) and detected by chemiluminescent-labeled antibody specific for DNA/RNA hybrids. Hybrid capture assays are available for C. trachomatis, N. gonorrhoeae, and human papillomavirus.

Nucleic Acid Amplification Tests

There are several methods for amplification (copying) of small numbers of molecules of nucleic acid to readily detectable levels. These NAATs include PCR, LCR, strand displacement amplification, and self-sustaining sequence replication. In each case, exponential amplification of a pathogen-specific DNA or RNA sequence depends on primers whose annealing to the target sequence leads to specific amplification of this sequence. The amplified nucleic acid can be detected after the reaction is complete or (in real-time detection) as amplification proceeds. The sensitivity of NAATs is often greater than that of other methods such as culture or nucleic acid probes. PCR, the first and still the most common NAAT, requires repeated heating of the DNA to separate the two complementary strands of the double helix, hybridization of a primer sequence to the appropriate target sequence, target amplification by PCR for complementary strand extension, and signal detection via a labeled probe. Methods for the monitoring of PCR after each amplification cycle—via either incorporation of fluorescent dyes into the DNA during primer extension or use of probes that fluoresce after binding a target sequence—have decreased the time required to detect a specific target. An alternative NAAT employs TMA, in which an RNA target sequence is converted to DNA, which then is exponentially transcribed into an RNA target. The advantage of this method is that only a single heating/annealing step is required for amplification.

Identification of otherwise difficult-to-identify bacteria involves initial PCR amplification of a region of 16S rDNA that is highly conserved within species and subsequent sequencing of the PCR product. Although 16S sequencing is not as rapid as other methods and is still relatively expensive for routine use in a clinical microbiology laboratory, it is becoming the definitive method for identification of unusual or difficult-to-cultivate organisms.

Next-Generation and Metagenomic Nucleic Acid Tests

Next-generation sequencing refers to high-throughput methods capable of rapidly sequencing large numbers of DNA molecules in a complex mixture of sequences. Next-generation sequencing has made it possible to rapidly determine the sequence of bacterial and viral genomes and to detect the nucleic acid of pathogens directly in some types of human samples. Whole-genome sequencing of bacteria can be used for species identification and potentially to predict antibiotic susceptibility; however, there are far faster and less expensive methods available for these applications. Whole-genome sequencing of bacteria for typing, which is increasingly practical and useful, is discussed below. The analysis of large numbers of DNA sequences present in complex mixtures of diverse DNA, called metagenomic sequencing, can be used to detect pathogens in some human tissues. Because high levels of human DNA make the acquisition and analysis of data difficult, this method has primarily been applied to CSF samples, which have relatively little human DNA compared to other sample types. There are several case reports in which metagenomic testing has been used to detect bacteria and viruses associated with CNS infections. A small number of laboratories offer metagenomic sequencing for diagnostic testing. Because the test is expensive and slow, targeted testing for the most likely pathogens should be done before metagenomic sequencing is considered.

Quantitative Nucleic Acid Tests

With the advent of newer therapeutic regimens for HIV-associated disease, cytomegalovirus infection, and hepatitis B and C virus infections, the response to therapy has been monitored by determining both genotype and viral load at various times after treatment initiation. Quantitative NAATs are available for HIV (PCR), cytomegalovirus (PCR), hepatitis B virus (PCR), and hepatitis C virus (PCR and TMA). Many laboratories have validated and now perform quantitative assays for these and other pathogens (e.g., Epstein-Barr virus), using analyte-specific reagents for NAATs.

Branched-chain DNA (bDNA) testing is an alternative to NAATs for quantitative nucleic acid testing. In such testing, bDNA attaches to a site different from the target-binding sequence of the original probe. Chemiluminescence-labeled oligonucleotides can then bind to multiple repeating sequences on the bDNA. The amplified bDNA signal is detected by chemiluminescence. bDNA assays for viral load of HIV, hepatitis B virus, and hepatitis C virus have been approved by the FDA. The advantage of bDNA assays over PCR is that only a single heating/annealing step is required to hybridize the target-binding probe to the target sequence for amplification.

Applications of Nucleic Acid Tests

The number of FDA-approved NAATS has increased dramatically and includes tests for Mycobacterium tuberculosis, N. gonorrhoeae, C. trachomatis, groups A and B Streptococcus, and methicillin-resistant S. aureus. FDA-approved multiplex NAAT panels for simultaneous detection of several respiratory, central nervous system, or gastrointestinal pathogens and limited antibiotic resistance markers also are available. Such panels are also available for positive blood culture broths. Although the use of these syndromic panels has increased in recent years, several issues regarding their use still need to be resolved, including what to do when more than one target is detected. Again, many laboratories have used commercially available reagents and analyte-specific reagents to create laboratory-developed tests for diagnostic use. Nucleic acid tests are useful for detection and identification of difficult-to-grow or noncultivable bacterial pathogens such as Legionella, Ehrlichia, Rickettsia, Babesia, Borrelia, and Tropheryma whipplei. In addition, methods for rapid detection of agents of public health concern, such as Franciscella tularensis, Bacillus anthracis, smallpox virus, and Yersinia pestis, have been developed and are available in regional (state) laboratories and at the Centers for Disease Control and Prevention.

Nucleic acid tests are also used to determine how close the relationship is among different isolates of the same species of pathogen. The demonstration that bacteria of a single clone have infected multiple patients in the context of a possible source of transmission (e.g., a health care provider) offers confirmatory evidence for an outbreak. Pulsed-field gel electrophoresis is an accessible and fast method for bacterial strain analysis. This method involves the use of restriction enzymes that recognize rare sequences of nucleotides to digest bacterial DNA, resulting in large DNA fragments. These fragments are separated by gel electrophoresis with variable polarity of the electrophoretic current and then are visualized. Similar band patterns (i.e., differences in ≤3 bands) suggest that different bacterial isolates are closely related, or clonal. Simpler methods of strain typing include sequencing of single or multiple genes and PCR-based amplification of repetitive DNA sequences in the bacterial chromosome. Whole bacterial genome sequencing has been used in investigation of many suspected outbreaks and is routinely used in public health laboratories to monitor for outbreaks of food-borne infection with Listeria monocytogenes. This method is likely to be increasingly used for typing of food-borne pathogens as it represents the new gold standard in bacterial strain analysis.

Future applications of nucleic acid testing probably will include the replacement of culture for identification of many pathogens by additional multiplex NAAT and solid-state DNA/RNA chip technology, in which thousands of unique nucleic acid sequences can be detected on a single silicon chip.

A principal responsibility of the clinical microbiology laboratory is to determine which antimicrobial agents inhibit a specific bacterial isolate. Such testing is used for patient care and for monitoring of infection control problems, such as methicillin-resistant S. aureus or vancomycin-resistant Enterococcus faecium. Two general approaches are useful. First, phenotypic testing is done, with exposure of the bacterial isolate to antibiotics by various methods and measurement of the effect on the growth of the organism. Second, genotypic testing is done to detect the presence of genes that correspond to resistance to specific antibiotics. Phenotypic testing can be done for a categorical response—usually susceptible, resistant, or intermediate. The “susceptible, dose-dependent” category indicates that the organism is susceptible but that dosing at the higher or more frequent recommended levels is needed. Testing for a categorical result can involve either the placement of paper disks containing antibiotics on an agar surface inoculated with the bacterial strain to be tested (the Kirby-Bauer or disk-diffusion method), with measurement of the zones of growth inhibition after incubation, or the use of broth cultures containing a set concentration of antibiotic (the breakpoint method). These methods have been calibrated carefully against quantitative methods and clinical experience with each antibiotic, and zones of inhibition and breakpoints have been determined on a species-by-species basis. Alternatively, phenotypic testing can be used to determine the lowest concentration of antibiotic that inhibits visible bacterial growth (the minimal inhibitory concentration or MIC).

After an MIC has been determined, the tubes in which no growth is seen can subcultured to determine the lowest concentration of antibiotic that reduces the number of viable bacteria by 99.9% (the minimal bactericidal concentration or MBC). The MIC value can be given a categorical interpretation of susceptible; susceptible, dose-dependent; resistant; or intermediate. Quantitative susceptibility testing using microbroth dilution in microwell plates or other miniaturized testing platforms has been automated and is used commonly in clinical laboratories. The epsilometer test (E-test), a novel method for determination of the MIC, uses a plastic strip with a known gradient of antibiotic concentrations along its length. When the strip is placed on the surface of an agar plate seeded with the bacterial strain to be tested, antibiotic diffuses into the medium, and bacterial growth is inhibited. For some organisms, such as obligate anaerobes and some β-hemolytic streptococci, routine susceptibility testing generally is not performed because of the difficulty of growing the organisms or the predictable sensitivity of most isolates to specific antibiotics.

Genotypic tests for antibiotic resistance are accurate when only a few genes are highly likely to cause resistance to specific antibiotics. Genotypic tests usually provide results 1 or 2 days earlier than phenotypic tests. There are FDA-approved genotypic tests for methicillin resistance in S. aureus, vancomycin resistance in Enterococcus species, and carbapenem resistance in enteric gram-negative bacilli (e.g., E. coli, Klebsiella species). These are available as stand-alone tests that can be performed on isolated bacterial growth or as part of multiplex assays that also detect specific pathogens in positive blood culture broths.

With the advent of many new agents for treating yeasts and systemic fungal infections, the need for testing of individual isolates for susceptibility to specific antifungal agents has increased. Susceptibility testing of yeasts is now performed in many laboratories. Current recommendations for interpretation of results with yeasts are specific for species rather than for genera, as in the past. The category “susceptible, dose-dependent” (defined above) is used for Candida glabrata with some MICs of fluconazole. Susceptibility testing with molds (fungi with hyphae), such as Aspergillus species, remains technically challenging and time consuming and is usually performed in commercial or reference laboratories. For most antifungals, an MIC is reported, but for echinocandins a minimum effective concentration (MEC) will be reported. The MEC is the lowest concentration that causes specific changes in the morphology of the fungus, and it is more reproducible than the MIC. Results of mold susceptibility testing are usually reported without conventional interpretations of susceptible, intermediate, or resistant. An epidemiologic cutoff value (ECV) may be reported for some results of antifungal susceptibility testing. Isolates with an MIC or MEC below the ECV lack genetic mutations that lead to increased MIC or MEC, but the ECV should not be used as the sole basis for the selection of therapy.

See Chaps. 185 and 186.