Nucleic acid amplification assays for microbial diagnosis: Challenges and opportunities

Abstract

The ability to detect small quantities of microbial nucleic acids by means of polymerase chain reaction (PCR) and other nucleic acid amplification methods has provided clinicians with a powerful new set of tools for the detection of low levels of microbial agents in clinical samples. However, as is often the case with new advances in microbiology, our ability to detect microorganisms has advanced faster than our ability to interpret the biologic meaning and medical consequences of microbial detection. In the past, new techniques for microbial detection have been evaluated by comparison with an established reference method. The term “gold standard” is often used to describe the reference method, although given the ups and downs of the price of precious metals during the past few years, this term should probably be discarded as anachronistic. In any case, the comparison of novel methods with established reference techniques has been used for the past century to define the sensitivity, specificity, and predictive values of the new diagnostic techniques. This approach has led to the rational introduction of many new diagnostic techniques into the clinical laboratory with the resulting improvement in the ability of clinicians to accurately diagnose infectious diseases. Although improved diagnostic methods have had an effect on all branches of medicine, the effect has been particularly important in pediatrics because of the high prevalence of infectious diseases in pediatric practice. The comparison of a newly developed diagnostic method is problematic when the new technique can detect the presence of the microbial pathogen in cases where it is undetectable by previously available methods. This is particularly the case for organisms such as Pneumocystis carinii, which cannot easily be grown in cell culture and for which direct detection methods must be used for organism identification. In such cases, the reference methods may not be ideal in terms of their sensitivity and specificity. It is clear that PCR can detect microbial genomes extracted from microorganisms in numbers that are below the detection limits of other assays. Howev