What Does a Polymerase Chain Reaction (PCR) Result Really Mean? The Dirty Secrets of Molecular Diagnostics
Bruce A. Rideout, DVM, PhD, DACVP
Department of Pathology, Center for the Reproduction of Endangered Species, Zoological Society of San Diego, San Diego, CA, USA
Molecular diagnostic tests are now routinely offered by most government, university, and private commercial laboratories. However, because these tests are relatively new, many clinicians and pathologists have limited experience with their selection and interpretation. In addition, these new diagnostic methods bring with them new pitfalls and therefore require new criteria for interpretation.3 The purpose of this presentation is to explain some of the advantages and disadvantages of molecular diagnostics, with the hope that it will be an aid to test selection and interpretation.
There are a number of different molecular diagnostic tools available today for diagnosis of genetic and infectious diseases, but this talk will focus on the use of the polymerase chain reaction (PCR) to detect specific nucleic acid (DNA or RNA) sequences for infectious disease diagnosis. This is by far the most common application of molecular diagnostics in the zoo and wildlife setting. The purpose of PCR is simply to amplify the target DNA to the levels necessary for detection. The amplified DNA can then be identified by its molecular weight (by comparing its migration distance on a gel to controls of known molecular weight) or by its base sequence (as determined by an automated DNA sequencing instrument, or by hybridizing a labeled DNA probe to the amplified DNA—called Southern blotting). PCR can also be used to identify RNA by a method called reverse transcriptase (RT)-PCR, which reverse transcribes the RNA into cDNA before amplification. This is followed by identification of the amplified cDNA by the methods described above. Although technically more complicated, this type of PCR is important because many viruses have RNA genomes.
PCR as a diagnostic tool has many advantages, including extremely high sensitivity, potentially high specificity, lack of host-species specificity (very important for zoo animals!), potentially short turnaround time, minimal sample size requirements, and the ability to rapidly develop new tests for unknown or unculturable agents.4-6 PCR can also be applied to many different sample types, including paraffin-embedded tissues, and is useful for molecular strain typing in epidemiologic investigations.5 Finally, molecular identification of pathogens is safer for laboratory personnel than in vitro cultivation, which requires more stringent biosafety procedures and special permitting for certain agents.6
Less well publicized are the disadvantages of PCR as a diagnostic tool. These include the fact that although the test itself is not host-species specific, validation of specific tests for diagnostic applications is! What this means is that a laboratory offering PCR for a specific agent in a specific host, such as feline herpesvirus in the domestic cat, has validated that test to reliably identify only that specific agent in that specific host. The validation process involves demonstrating that there are no other common infectious agents, commensal organisms, or host DNA segments that would be accidentally amplified by the PCR primers, thus producing a false positive. This type of validation enables a positive test to be based solely on finding a band of the appropriate molecular weight on a gel, thus foregoing the labor-intensive sequencing or Southern blotting required to more definitively identify the PCR product. While this approach is appropriate when it is applied to the agent and host species for which it was validated, using domestic animal tests in zoo animals can result in unpredictable false positives or false negatives. In other words, in most of the testing situations we face, finding a band of the correct molecular weight on a gel is not sufficient to call the test positive. There are many ways in which PCR can accidentally amplify unrelated sequences to create a band of the correct molecular weight, but the wrong sequence. It is important when using commercial molecular diagnostics in zoo animals to have a plan for confirming any positive bands, either by direct sequencing, Southern blotting, or by another confirmatory test.
Additional disadvantages of PCR include the fact that it is not usually quantitative, which can make clinical significance of a result more difficult to interpret. For example, the quantitative nature of conventional tests is often useful in establishing the clinical significance of the test result, as when a clinically ill animal has a rising or very high titer to a likely etiologic agent, or a bacterial culture yields heavy growth of a pathogen. In contrast, a PCR result may be reported as positive even though only a few individual organisms or viral particles are present. In many circumstances, infectious agents present in such low numbers could be clinically insignificant. Such “clinical false positives” can also occur because PCR doesn’t distinguish between viable or dead microorganisms. Other disadvantages of molecular diagnostics are the technical difficulty of the procedures, the expensive equipment required, the high risk of cross contamination of amplified DNA in the laboratory leading to false positives, and the contribution of sampling error to the occurrence of false negatives. Finally, sample handling requirements can be complicated. RT-PCR usually requires very fresh samples and special transport media to prevent degradation of the RNA, which is much more labile than DNA. It is always important to scrupulously avoid contamination of samples (e.g., microtome blades must be changed between every block when cutting paraffin sections for PCR). Certain sample types also carry inherent limitations, for example PCR on paraffin embedded tissues is generally limited to identifying base sequences of less than 500 bases, and direct PCR on fecal samples often results in false negatives due to the presence of unidentified inhibitors.
So how can a busy clinician make the most of molecular diagnostics?
1. First, be aware that the dirty secret of molecular diagnostics is that a “positive” test may only mean a band of a particular molecular weight on a gel. It doesn’t necessarily mean that that band corresponds to the specific infectious agent you are looking for.
2. Know what your result means (i.e., try to get a feel for potential sources of false positives or negatives and have a plan to confirm any positives, either by direct sequencing, Southern blotting, or alternative testing).
3. Know what you are asking for and why. Is the point of the test to determine whether a particular agent is present, to confirm results of another test (e.g., serology), or to characterize and more specifically identify an agent already known to be present? Be aware that test performance will generally be much lower when you are screening healthy animals for a disease of low prevalence as compared to testing clinically ill animals for a likely etiologic agent.
4. Understand how to interpret the clinical significance of a positive or negative result. For example, does the agent cause latent infections or asymptomatic carrier states? If so, a positive PCR result might not be any more informative than positive serology. Does the presence of the agent alone have clinical significance or is quantitation necessary to establish significance? Could a negative result be due to sampling error?
Molecular diagnostic technology is still changing rapidly, with many advances on the horizon. DNA-Chip technology has the potential to provide rapid molecular diagnostics for many different pathogens in one simple on-site test.1 Real-time quantitative PCR is available now and can provide very rapid results with specific confirmation of positives incorporated into the test.2 Other applications include simultaneous bacterial identification and antibiotic susceptibility testing within hours, searching for residual neoplasia following chemotherapy, monitoring response to treatment for infectious diseases, and representational difference analysis for identification of any unknown agents in a sample.5,6
1. Aitman, T.J. 2001. DNA microarrays in medical practice. BMJ. 323:611–5.
2. Foy, C.A. and H.C. Parkes. 2001. Emerging homogeneous DNA-based technologies in the clinical laboratory. Clin. Chem. 47:990–1000.
3. Fredericks, D.N. and D.A. Reiman. 1996. Sequence-based identification of microbial pathogens: a reconsideration of Koch’s postulates. Clin. Microbiol. Rev. 9:18–33.
4. Gao, S.J. and P.S. Moore. 1996. Molecular approaches to the identification of unculturable infectious agents. Emerg. Infect. Dis. 2:159–167.
5. Gilbert, G.L. 2002. Molecular diagnostics in infectious diseases and public health microbiology: cottage industry to postgenomics. Trends Mol. Med. 8:280–287.
6. Tang, Y.W., G.W. Procop, and D.H. Persing. 1997. Molecular diagnostics of infectious diseases. Clin. Chem. 43:2021–2138.