So happy together: Integrating molecular and serological testing

By: John Brunstein   

In the days before molecular diagnostics (MDx) for infectious diseases, two general classes of methodology were the standard: culture (plate media for bacteria and fungi, cell culture for viruses), and serological testing (either based on antibodies against the suspected pathogen, thereby directly testing for its presence; or looking to detect endogenous antibodies, as evidence of near- or long-term prior infection). While MDx methods have a number of strengths compared to these older approaches (small sample sizes, extreme sensitivity, excellent specificity, easy automation, rapid test development path), they remain in widespread use because each still has particular advantages over MDx. Culture, for example, has the ability to look for unexpected or even novel pathogens, while MDx’s specificity means that in the formats commonly used it can only be used to query for known, specific targets. (Whole sample deep sequencing, of course, would be an exception to this, but it’s currently far more expensive and labor-intensive than culturing approaches).

Serologic tests, either for detection of pathogen or of pathogen-specific antibodies in a patient sample, can be very inexpensive, rapid, and simple to perform and often don’t even require complex instrumentation. While virus culture capabilities have been done away with in many labs, most still have serologic testing capabilities. In this month’s column, we’re going to examine some of the ways in which this can be used to complement MDx methods.

The combined approach

A first point of observation is that sometimes, too much sensitivity is a bad thing. A polymerase chain reaction (PCR) assay for a pathogen with capacity to detect one organism’s worth of genetic material or less in a sample may actually be misinformative. (If the idea of detection of less than one pathogen genome’s worth of material has you puzzled, keep in mind that some pathogens contain multicopy repeat elements, and MDx assays optimized for sensitivity are often directed against these, allowing for sub-single-genome limits of detection.) Detection of pathogen DNA or RNA isn’t proof of organism viability and can arise weeks, months, or in some cases even years after an infection has been cleared. In a case such as this, a positive PCR result can mislead by providing an apparent but erroneous root cause for a clinical presentation, leading to a lack of further diagnostic investigation and a lack of appropriate treatment for the actual cause. The risks of this are perhaps lessened if a quantitative PCR method is used as opposed to a purely qualitative assay; very low target signals compared to what’s seen in an ongoing primary infection context might be a warning sign to consider.

For comparison to this MDx-only approach, let’s consider what combinations of results (and likely interpretations) we might observe if samples were tested in parallel by target-specific PCR and by serologic testing for anti-target antibodies with class subtyping. For sake of this example, we’ll assume the use of a qualitative PCR. Our results and likely interpretations matrix would look something like Table 1 (Neg = negative result, Pos = positive result).

While these interpretations are of necessity simplistic (they assume for instance normal immunological response in the patient), they serve to illustrate how either MDx or serologic testing on its own might have been misleading. A PCR positive arising from trace residual nucleic acids of a long-since cleared infection, or a serologic-test negative in the immediate earliest stage post-infection, are the two most likely scenarios where an integrated combination of these test strategies would be of obvious benefit in getting accurate diagnoses. The scenario of an active infection by an agent with altered and now non-detected genetic target sequence, while much less likely, is also a possibility, and it’s certainly a benefit that integrated testing would have potential to detect this.

Table 1
Table 1

Case study: HHV-6

While the above discussion has been hypothetical and general in nature, no discussion of the utility of pairing MDx with serological methods would be complete without consideration of one widely recognized specific example—that of HHV-6 (human herpesvirus 6) testing. For those of you who may not be up to date on your viral infections, HHV-6 comes in two genetic variants, HHV-6A and HHV-6B, with 6B being the causative agent of exanthema subitum (more commonly known as roseola). Some estimates have suggested that nearly one in five (20 percent) of infant emergency room visits for fever are attributable to this virus. HHV-6 has been additionally implicated or found in association (with unknown significance) in a number of other serious medical conditions, including MS and female infertility.

Herpes viruses are double-stranded DNA viruses with relatively stable genomes, making them generally good targets for MDx-based assays. In the case of HHV-6, though, there’s an additional complication: it’s known to sometimes integrate (that is, splice itself into the host chromosome) after infection in immunocompetent hosts. Such integrated HHV-6 DNA, while not necessarily biologically active, will, of course, lead to a positive PCR result. More important, it can also be vertically transmitted, meaning a child born to a parent with a germline integrant HHV-6 copy will carry a congenitally acquired HHV-6 copy in every cell. This is not as rare an occurrence as might be expected, with one study of UK blood donors providing an estimated incidence of about one percent.1 The perhaps counterintuitive outcome of this is that when quantitative PCR methods are used, particularly high HHV-6 viral loads (above 5.5 log10 per ml in peripheral blood) are more likely to be indicative of a non-active integrant virus than a primary infection.2

On their own, then, qualitative HHV-6 PCR assays can be potentially misleading. If a child presents with acute febrile illness and HHV-6 PCR tests positive and is taken as the causative agent, precluding further testing, it’s quite possible that a true etiologic agent may be overlooked and corresponding appropriate therapy missed. One possible approach to avoiding this is to use an RT-PCR- based assay, which would be specific for HHV-6 RNA and thus provide definitive evidence of viral transcription and thus (likely) replication. Another approach is that which is our focus today: utilizing a paired serology and molecular result. PCR coupled with screening for appropriate serological markers for HHV-6 infection represents a cost-effective and rapid method for increasing diagnostic specificity to the context of a primary infection.

While serological methods and MDX methods are frequently used in exclusion to each other to good effect, it’s thus worth keeping in mind that they query fundamentally different biological markers for infection. In some cases, they can be applied together to arrive at the most accurate diagnostic conclusion.


  1. Leong HN, Tuke PW, Tedder RS, et al. The prevalence of chromosomally integrated human herpesvirus 6 genomes in the blood of UK blood donors. Journal of Medical Virology. 2007;79(1)45-51.
  2. Pellett PE, Ablashi DE, Ambrose PE, et al. Chromosomally integrated human herpesvirus 6: questions and answers. Reviews in Medical Virology. 2012;22(3):144-155.



John Brunstein, PhD, is a member of the MLO Editorial Advisory Board. He serves as President and Chief Science Officer for British Columbia-based PathoID, Inc., which provides consulting for development and validation of molecular assays.

So happy together: Integrating molecular and serological testing
John Brunstein
John Brunstein, PhD, is a member of the MLO Editorial Advisory Board. He serves as President and Chief Science Officer for British Columbia-based PathoID, Inc., which provides consulting for development and validation of molecular assays