Point-of-Care Testing for Influenza
At some point in the coming months, you'll almost certainly hear that a friend, family member or coworker "got the flu." Of course, some cases will actually result from an aversion to going to work on a Monday, while others will indeed be caused by the influenza virus. The remainder will be illnesses due to one of a multitude of other viruses or bacteria that can cause similar symptoms to influenza. Unfortunately, diagnostics for the first situation are not yet mature. However, to address the latter two cases, we have recently seen a proliferation of FDA-approved point-of-care tests (POCT) for detection of viral antigens or nucleic acid in nasal secretions, nasopharyngeal aspirates, and/or other respiratory specimen types.
Influenza Virus POC Testing Options
In addition to tests that detect surface antigens, there is a relatively new nucleic acid amplification-based test (NAAT) that can be performed at the point of care. NAAT testing uses nucleic acid probes that bind specifically to a region of the influenza genome. Amplification generates a fluorescence signal that is read by a dedicated instrument. NAAT tests used to be restricted to the clinical laboratory because of the complexity of test setup. These tests are read using an automated instrument but require more expensive reagents and, of course, investment in the instrument itself. By definition, POCT are not performed in a central laboratory setting. Therefore, the ideal POCT must be simple to perform by individuals providing patient care, and have low probability of yielding incorrect results. Tests meeting these criteria are termed CLIA-waived and can be performed by personnel directly caring for patients. Sites performing waived testing do need a CLIA certificate but are exempt from other CLIA requirements. There are currently 8 CLIA-waived influenza POCT. The most commonly used of these are rapid influenza diagnostic tests (RIDTs) that use monoclonal antibodies on the test instrument to detect viral nucleoprotein antigens from the patient in an immunochromotographic format. RIDTs provide qualitative (positive vs. negative) results, are simple to perform, and the majority do not require specialized instrumentation for setup or interpretation.
Is It the Flu? Influenza POC Caveats
When I visit the doctor for a potential respiratory infection, I ask, "Is it the flu?" I don't ask, "Do my nasal secretions have influenza antigens or nucleic acids above the limit of detection of your chosen test kit?," despite the fact that this is the only question answerable by diagnostic tests. Luckily, both RIDT and NAAT-based POCT typically have good (>98%) specificity, so answering "yes" to the second question is usually a good indication that it is indeed the flu. As such, positive influenza POCT are informative and may reduce unnecessary antibiotic prescriptions and allow for early identification of influenza outbreaks.
However, influenza antigen POCT are notorious for low sensitivity, which according to a recent meta-analysis, was only 62.3% for RIDTs. The low sensitivity of these tests likely has both technical and biological underpinnings. One technical challenge is the requirement for a very wide dynamic range in test readouts due to the fact that viral loads vary by >2 logs among patients early in infection, which is the period when antiviral medications are most effective and, therefore, when POC diagnostics are most likely to be useful.
A major biological factor that contributes to accuracy is the high mutability of the influenza virus itself. Specifically, influenza RNA polymerase is notoriously error-prone even in comparison to other viruses. Mutations arise at a rate of approximately 1 x 10-6 per site during every cycle of infection, corresponding to about one error in every virus particle. To put this in perspective, Escherichia coli, a bacterium that is readily mutated in the laboratory, has an approximately 10,000-fold lower mutation frequency because of the proofreading activity of DNA polymerases. High mutation rates result in subtle structural changes to viral proteins, a phenomenon called antigenic drift (see schematic, right).
Further complicating matters, the influenza genome is segmented into 8 strands, and influenza isn't very picky about how these segments are placed into viral particles. Therefore, coinfection with different influenza strains can result in a hybrid virus with segments from two (or even more) viral types. This can create a virus with a novel array of antigens essentially instantaneously, an event termed antigenic shift.
Antigenic drift and shift benefit influenza by allowing it to evade the immune system. From a diagnostic perspective, these mutations lead to structural changes in antigenic proteins which may impact the ability of RIDTs to recognize specific epitopes. NAAT tests are likely even more sensitive to antigenic drift. Small genetic changes occurring during an influenza season, including those which do not alter the structure of antigenic proteins, can preclude probe binding and significantly impact NAAT sensitivity.
Quality Assurance of Influenza POCT
Despite the simplicity of POCT, quality assurance and operator training/competency programs for these tests should not be overlooked. These programs are of particular importance for influenza POCT because the most commonly used products are read manually, making variability in test reading possible among different operators. This also means that test results are not automatically uploaded to a laboratory information system (LIS) or electronic medical record (EMR), demanding specific procedures to ensure consistent, accurate documentation of test results.
In addition to the technical aspects of testing and reporting results, the clinical implications of influenza POCT must also be considered. Owing to their relatively low sensitivity compared to other types of tests, CDC guidance recommends that "antiviral treatment should not be withheld from patients with suspected influenza, even if they test negative" to account for the low negative predictive value of the most common POCT. Nevertheless, from a patient's point of view, it is easy to conflate a negative test result with "it's not the flu." Therefore, health care providers should interpret negative results with caution and also be ready to educate patients about test limitations.
The Role of the Clinical Laboratory in Influenza POCT
Follow-up testing by the clinical laboratory using more sensitive tests such as immunostaining or RT-PCR should always be considered when the test result is unexpected in light of the prevalence of influenza in the community (for example, a negative test when prevalence is very high). Additionally, POCT do not have the ability to detect novel influenza viruses or viral subtypes. As such, CDC suggests that negative tests should be confirmed by the clinical laboratory when POC testing is used for outbreak monitoring.
Accurate diagnosis of influenza is important for several reasons, including individual patient care, infection control and epidemiological monitoring. However, until recently, POCT for influenza have suffered from a lack of sensitivity. New FDA guidelines have recognized this sensitivity defect and reclassified influenza POCT as class II devices, subjecting these tests to special controls, including a requirement for >80% sensitivity versus RT-PCR.
There is not yet sufficient data to determine whether this regulatory change will result in the introduction of more sensitive influenza POCT. However, we can be confident that antigenic shift and drift will continue to challenge POCT sensitivity. Therefore, the clinical laboratory must always be ready to support POCT through follow-up testing.
The above represents the opinions of Dr. K.P. Smith and not necessarily of the American Society for Microbiology.