False-negative results occur when the analyte is present in the specimen but the molecular assay result is negative. In quantitative molecular assays, aberrant results may include completely negative results or inaccurate quantitation. For example, the assay may report viral loads below the threshold of detection when there is virus present in the specimen. There are several conditions in which false-negative results can be obtained in molecular diagnostics. Interferences that result in false negatives or inaccurate quantitation can occur in pre-analytical or analytical phases of testing.
Inhibition in Amplification Assays
Inhibition of amplification (complete or partial fail- ure to generate amplified DNA or RNA because of inhi- bition of enzymes used in amplification reactions) can be a cause of negative results or inaccurate quantitation [35]. The presence of inhibitors can cause false-negative results in qualitative assays or can reduce amplification efficiency, causing inaccurate results in quantitative assays. Inhibition of amplification reactions is a well- recognized phenomenon in the diagnostic laboratory, but the reasons for inhibition are not always well understood [36]. Many substances from a variety of sources can inhibit PCR reactions. Inhibitors can be present in the clinical specimen (e.g., hemoglobin and 331
INTERFERENCES IN MOLECULAR DIAGNOSTICS: FALSE-NEGATIVE RESULTS IN MOLECULAR ASSAYS
lactoferrin from blood specimens) or be carried over from the transport device (e.g., heparin from a blood collection tube) or from the post-collection procedures such as nucleic acid extraction[35].
Some inhibitors are well characterized and the mechanisms by which they function are understood.
However, for some specimens and assays, inhibition may occur but the inhibitory substances are not defined and/or the mechanism of inhibition is not understood.
Inhibitory substances may bind directly to the target nucleic acids, blocking enzyme binding, or they may bind directly to amplification enzymes, preventing enzymes from binding the target. Although inhibition is a well-recognized phenomenon, it is not always pos- sible to determine exactly what the problem is when inhibition occurs. Inhibition can affect different amplifi- cation reactions differently and to different levels [35,36]. Inhibitors are especially problematic in quanti- tative assays because they may cause aberrant results (inaccurate quantitation) or they may inhibit amplifica- tion altogether. Commercially available quantitative PCR assays use software algorithms that normalize the signal from the internal calibrators present in each reac- tion vessel. Patient test results are based on normalized values; if partial or complete inhibition has occurred, it is detected by the system and interpreted correctly.
Laboratories that develop in-house quantitative PCR should implement procedures using normalization to ensure accurate results in their quantitative PCR.
Mechanisms by which Inhibitors Function
There are several mechanisms by which inhibitors function, usually targeting the enzymes involved in amplification or interfering with binding of the enzyme to the target nucleic acid:
1. Blocking active site of enzymes: Some substances block active sites of enzymes used in amplification reactions. Heme compounds from blood function to inhibit Taq polymerase in this manner.
2. Inactivating enzymes: Inhibitors can also act by denaturing or degrading the enzymes used in amplification (e.g., if proteinases are present in the sample).
3. Binding target nucleic acids, preventing the target from being accessible to primers or enzymes.
4. Binding divalent cations: Inhibition can occur when divalent cations (Mg21) necessary as co-factors for polymerase enzymes are bound by EDTA.
Commonly Encountered Inhibitors and their Sources
Inhibitors are present in clinical specimens, collec- tion and transport devices, and can be introduced
during processing or nucleic acid extractions (Table 21.1). Amplification tests that are performed directly on specimens without a separate nucleic acid purification step are especially prone to inhibition.
INHIBITORS IN SPECIMENS
Many clinical specimens, such as urine, feces, blood, and sputum, are complex and contain inhibitory substances. Not all inhibitors have been characterized.
Specimens collected in preservatives or fixed in alco- hols or formaldehyde solutions are especially subject to inhibition during PCR, even when nucleic acids have been extracted and purified prior to use in the assay. Some inhibitors have been well-characterized.
Hemoglobin is a potent inhibitor of PCR amplification and a well-characterized inhibitor of the Taq polymer- ase [37]. Complex specimens such as stool, sputum, and urine usually contain inhibitors; assays for microbes from complex specimens are optimally per- formed on extracted nucleic acids to reduce the chances of co-purification of inhibitors.
INHIBITORS IN TRANSPORT DEVICES
Specimen collection devices such as blood collection tubes may contain substances that function as amplifi- cation inhibitors, including heparin and EDTA. EDTA can potentially inhibit PCR by binding divalent cations that are used as co-factors by amplification enzymes [35]. Heparin is easily co-purified with DNA and a strong inhibitor of PCR at levels as low as 0.032 U/mL in the PCR reaction[38]. If these substances are carried through the nucleic acid extraction procedure, inhibi- tion of amplification reactions may occur.
TABLE 21.1 Common PCR Inhibitors
General Source Inhibitory Substance
Specimens Blood—hemoglobin, IgG, lactoferrin Blood—DNA binding proteins Muscles—myoglobin
Sputuma
Stoola—bile salts, complex polysaccharides
Tissues—melanin Urinea—urea Specimen collection tubes EDTA
Heparin
Nucleic acid extraction Detergents (Sarkosyl, SDS) Proteinases
Protein disrupting agents Guanidinium salts Phenol
Alcohols
aFor many complex clinical specimens, inhibitors have not been well-characterized.
INHIBITORS IN EXTRACTION REAGENTS
Nucleic acid extraction procedures can be sources of potent amplification inhibitors. Detergents used to lyse cells (e.g., Sarkosyl and sodium dodecyl sulfate (SDS)), protein disrupting agents (guanidine), enzymes used to digest cellular proteins (proteinases), organic extrac- tion compounds (phenol and chloroform), and alcohols may be inadvertently carried over from the extraction procedure into the purified nucleic acid samples. If present in high enough concentrations, these sub- stances can degrade polymerases or physically block polymerase from copying target nucleic acids[35].
Co-Purification of Inhibitors
Extraction and purification of nucleic acids prior to use in amplification procedures can usually eliminate inhibitors, unless they co-purify with the nucleic acid.
Extraction methods differ in their ability to remove inhibitors from specimens. Different extraction and purification methods may need to be tried if a labora- tory is developing its own assays. For example, melanin present in pigmented cells from skin and certain tumors is a potent inhibitor of PCR and usually co-purifies with standard DNA or RNA extraction procedures [39].
Addition of high concentrations of proteins such as bovine serum albumin can reverse this inhibition [40].
The nucleic acid extraction and purification procedures may be a source of inhibitors if residual materials from the extraction remain in the final sample.
Monitoring for Amplification Inhibition
The best way to determine if inhibitors are present in a specimen is to include an internal control in every reaction. Monitoring for inhibition is a requirement for laboratories that are performing diagnostic assays using molecular methods. Commercially available, FDA-approved diagnostic assays include internal amplification controls to monitor for inhibitors. Several different types of amplification controls can be used, and both internal and external control materials should be included in amplification assays. Internal controls are used to monitor the entire process from extraction to assay interpretation[41,42].
For infectious disease assays, internal controls can be designed so that the same primer set is used with a dif- ferent probe sequence. For real-time PCR systems, the tests should be designed to contain an internal amplifi- cation control in every reaction. The amplification plot can be used to determine if inhibition has occurred. In systems in which the amplification can be monitored using software, amplification curves can be examined to determine if inhibition is occurring. Inhibition plots can be prepared and monitored. Laboratories that are developing their own assays must design a method for
monitoring inhibition. Guidelines for choosing internal control materials are available [41,42]. If the result of the internal control is negative, the report can be modi- fied to indicate that the specimen was inhibitory rather than negative. For genetic assays, an unrelated target sequence can be amplified and detected along with the gene of interest. For example, an assay for a housekeep- ing gene such as β-actin can be performed simulta- neously with the target sequence.
Monitoring of the rate of inhibitors can be per- formed as part of the laboratory’s quality assurance program. Unusually high rates of inhibition can be an indication that a problem exists somewhere in the sys- tem. For genetic assays, the extracted DNA is usually quantified so the amount of DNA that is put into the reaction can be standardized. If UV spectroscopy is used, the purity of the extracted nucleic acids can be verified by checking the A260/A280 ratio. Low ratios (for DNA, a ratio of 1.8 is ideal) indicate contamination with protein or other substances. If the ratio is poor, the DNA can be re-extracted and the test re-run.
Strategies to Prevent Inhibition
Laboratories that are developing their own nucleic acid amplification methods should optimize all phases of testing (specimen selection, transport devices, and nucleic acid extraction/purification) to reduce the presence of inhibitors in the samples being analyzed.
Amplification assays should be developed with inclu- sion of substances that can facilitate amplification, such as bovine serum albumin. If the laboratory is monitoring for inhibitors using internal controls and detects an increased rate of inhibition, the assay can be re-optimized. Commercially available PCR assays now contain internal controls; inhibition of the internal con- trol can be monitored and if significant inhibition occurs, the test may be reported as being inhibitory rather than negative. Using additives such as bovine serum albumin can alleviate or reverse the effects of inhibitors in PCR. For home-brew assays, laboratories should optimize the assay prior to routine use of any additives.
One of the best methods to prevent inhibition is to ensure that highly pure nucleic acids are used on the assay. Nucleic acid extraction/purification methods that are based on the Boom technology of silica bind- ing [32] result in good-quality nucleic acids. The silica binding methods are widely available, in many for- mats, and usually result in highly pure nucleic acids.
Laboratories may wish to avoid the use of assays that do not incorporate a nucleic acid extraction and purifi- cation step, especially for specimens that are known to contain inhibitors such as blood or urine.
For assays performed on plasma, guidelines for specimen collection should be followed carefully.
333
INTERFERENCES IN MOLECULAR DIAGNOSTICS: FALSE-NEGATIVE RESULTS IN MOLECULAR ASSAYS
Drawing the correct amount of blood into the tube will ensure that the EDTA is diluted to a level that will not inhibit downstream PCR testing after extraction of nucleic acids.
Inhibition of PCR can often be alleviated by the use of substances that can facilitate amplification.
Laboratories that develop their own test methods can experiment with incorporating substances into the assay during the development phase. Substances that have been successfully used to relieve inhibition include bovine serum albumin[43]and DMSO[35].
Poor Quality of Target Nucleic Acid: Target Degradation
The quality of the target nucleic acid has a signifi- cant effect on the performance of the molecular assay.
Good-quality, intact DNA or RNA is needed for optimal detection. Amplification methods generally require purified nucleic acids that have been extracted through a nucleic acid extraction and purification pro- cedure. Several commercially available assays do not incorporate a nucleic acid extraction step prior to use in the assay; these assays are susceptible to inhibition.
Optimally, the nucleic acid extraction method will give high-quality, intact nucleic acids that are free from co- purified material or from residual reagents carried over from the purification process.
Fixed Tissues or Cells
Fixed tissues or cells present a challenge for obtain- ing good-quality target nucleic acids. Depending on the condition of the tissues, and the fixative that has been used, extraction protocols may require some form of pretreatment step (e.g., to dissolve paraffin if tissues are embedded or to remove other chemicals such as preservatives or fixing agents). Pretreatment protocols can utilize heat, organic chemicals such as xylene, or enzyme digestion. Pretreatments can reduce the amount or quality of target nucleic acid and can also result in inhibition of amplification if reagents are co- extracted with the target material. Laboratories should include positive and negative controls when possible, which are handled in the same manner as clinical specimens.
Degradation of the DNA or RNA Target
Nucleic acids can become degraded during transport or storage. RNA is especially prone to degradation, and special procedures must be implemented in the labora- tory when working with RNA. Degradation of target nucleic acids can result in false-negative results or in inaccurate quantification in quantitative assays [44].
Nucleases may be present in the specimens, and nucleic
acid may be degraded by nucleases. Nucleases may be removed by using a nucleic acid extraction procedure that degrades proteins. RNA is especially prone to deg- radation by RNases, enzymes that are ubiquitous in the environment. Target degradation can be prevented or minimized by maintaining the samples in the correct environmental conditions prior to use in the assay. All samples should be handled using procedures that are optimized to prevent degradation of nucleic acids. If degradation of the target is suspected, the quality of the extracted nucleic acid should be checked with UV spec- troscopy or some other method.
Sequence Mismatch between Primer and Target DNA: Role of Genetic Variation
Nucleotide sequence variation in the target DNA or RNA sequence can result in false-negative test results, for example, if there are mismatches between the sequences of the primers and the target DNA or if a gene deletion occurs. Microorganisms, especially viruses, are prone to this phenomenon because they can evolve rapidly. Mutations in the primer binding area can pre- vent amplification altogether or reduce efficiency of amplification.
Sequence Variation in Primer Binding or Probe Regions
A mismatch between the primers or the probe sequence can result in false-negative assay results or can give inaccurate quantitative results in viral load assays. Viruses are especially prone to this phenome- non because they evolve rapidly. Sequence variation should always be a consideration when designing assays, especially home-brew assays. HIV and HCV viral load assays are subject to this phenomenon. Both of these viruses exist as viral populations with signifi- cant genetic variation. Although quantitative assays for these viruses are usually targeted to conserved areas of the genome, there is enough genetic variation in the population such that occasional strains of these viruses may be missed altogether or underquantitated [45]. False-negative molecular assays for bacterial pathogens can also occur because of sequence varia- tion. PCR-based assays forN. gonorrhoeaemay be nega- tive when strains of the organism with genetic variation in the primer binding region are present in the specimen[46].
CASE REPORT: FALSE-NEGATIVE RESULT OF PCR TESTING FOR NEISSERIA MENINGITIDIS A 2-year- old boy presented to the hospital with meningitis;
empiric antimicrobial therapy had been started prior to obtaining CSF and blood specimens for culture.
A Gram stain performed on the CSF revealed Gram- negative diplococci; culture was negative. Real-time PCR testing for N. meningitidis using primers for the ctrA gene was negative. Because the Gram stain was consistent with N. meningitidis, a second PCR assay using 16 S rDNA primers was performed with positive results for N. meningitidis. During the time the testing was being performed, blood cultures grewN. meningi- tidis. The child recovered. Follow-up testing of the isolate from the blood culture showed sequence varia- tion in the ctrA gene, with polymorphisms present on one of the primer binding sequences, and a single nucleotide substitution in the probe binding area[47].
For laboratories performing in-house developed tests, primer and probe sequences should be verified. This can be done by searching publicly available databases.
Analyte below Limit of Detection of Assay
One of the most common reasons for false-negative results, especially in qualitative infectious disease assays, is the presence of the analyte at very low levels—below the analytical sensitivity of the assay.
Although nucleic acid amplification methods can theo- retically detect a single molecule of target, in reality this is rarely achieved. The analytical sensitivity of molecular amplification methods should be considered when the assay result is negative.
Pre-Analytical Considerations that can Influence Nucleic Acid Amplification Assay Results
The timing of specimen collection is important when using molecular assays for diagnosis of infec- tious diseases. The specimens must be collected during the acute phase of illness; often, the pathogen may be undetectable by the time the patient presents for medical care. Diagnostic methods for infections due to arthropod-borne viruses such as West Nile virus are especially subject to this type of error. Viremias (virus in the blood) are transient, and with rare exceptions, testing of blood for virus by PCR is likely to yield neg- ative results. A similar phenomenon occurs for CSF. In most cases, the recommended diagnostic strategy for arthropod-borne viruses such as West Nile will include serological testing in addition to PCR to avoid misdi- agnosis due to false-negative PCR testing.
The timing of specimen collection is critical for many assays, especially viruses, in which the analyte may be present transiently. Collecting a specimen dur- ing the acute phase of illness is highly recommended for respiratory viruses, arthropod-borne viruses, and many others. For respiratory viral assays, the optimal specimen type is usually a nasopharyngeal swab or
aspirate, and throat swabs are usually discouraged by the laboratory.
For some assays, concentration of the specimen prior to testing in PCR assays is a useful method for increasing the sensitivity of the assay. For example, when testing CSF for infectious pathogens such as her- pes simplex virus, concentration of the analyte via nucleic acid extraction may help increase the chances of pathogen detection.
Technical Problems with Assay
Pre-Analytical Considerations
Specimen collection, processing, and storage condi- tions are critical steps in molecular testing, and appro- priate conditions must be used for successful results.
For viral load assays and for laboratory-developed assays, parameters must be worked out systematically.
These usually include optimizing the concentration of nucleic acid to include in the assay, optimizing concen- trations of cations that amplification enzymes need, and optimizing reaction temperatures and ramp speeds. Occasionally, amplification assays exhibit fail- ure or reduced performance even when parameters are optimized and the assays have been functioning well in the laboratory. In these cases, thermal cycler func- tion should be checked. Even slight variations in annealing and extension temperatures can have detri- mental effects on reactions.