Bacteria or virus? That is the question
Volume 8, Issue 7 | July 26, 2023
In This Issue
Another promising simple blood test for Alzheimer’s
Genetic sequencing comes for the dogs
WGS for diagnosis can be interpreter-dependent
WGS for screening may not be ready for prime time
New and Noteworthy
Bacterial or viral infection? Two new POC rapid tests could help point the way
Acute respiratory infections (ARIs) drive 150 million US outpatient visits each year. The first step in ARI diagnosis is to determine whether the cause is viral or bacterial. Not easy to do as the symptoms of each are often clinically indistinguishable.
Many single-analyte tests that could aid diagnosis have been available for some time, including tests for C reactive protein (CRP), procalcitonin, and many other individual cytokines. However, they are perceived as not accurate enough to be routinely used, so physicians prescribe antibiotics “just to be on the safe side”. This practice is often useless (50 - 60% of the time as the culprit is a viral infection) and dangerous (breeding antibiotic resistance).
Two point-of-care tests that received FDA approval this month improve accuracy by combining analytes: Lumos Diagnostics’ cassette FebriDx (fingerstick blood – tests for host immune proteins CRP and MxA) and MeMed’s desktop device (whole blood – tests for host immune proteins CRP, TRAIL and IP-10). FebriDx claims 90% of viral positives and 59% of bacterial positives are accurate. MeMed uses a scoring system to assist physicians in evaluating patients on a more probabilistic basis.
Commentary: We see two challenges here. #1: Getting physicians to use these tests in the clinic. #2: Getting patients to agree to the test and accept the answer, “It’s a virus. No antibiotics for you.” Even the single-analyte tests would have significantly improved treatment accuracy despite their shortcomings, but writing an Rx has long been perceived as the quickest, cheapest, and safest path - which is a big part of the reason why antibiotic resistance has become such a problem.
Finger-prick blood test promising for Alzheimer's diagnosis, monitoring
As we’ve discussed in the last few weeks’ issues, we expect to see an explosion of new diagnostic tests for Alzheimer’s disease (AD). The Alzheimer's Association International Conference (AAIC) 2023 just concluded, so we are now hearing about these new technologies. Most transformational will be those that are both accurate and as non-invasive as possible.
An exciting development from the conference that has the potential to tick both those boxes: an AD biomarker test based on blood from just a finger-prick on a dried blood spot card. What’s so exciting? The card can be transported at room temperature. If this research holds up to large clinical trials in comparison to cerebrospinal fluid, it could be a logistically easy (and potentially cost-effective) way to diagnose AD.
Didn’t pick up Fifi’s poop? Her genes will lead les gendarmes to you
We’ve been writing a lot about genetic sequencing recently, but here’s a use case we hadn’t considered. Dog owners in the southern French town of Béziers will be required to have their dogs genetically tested - so that poop left on the streets can be matched to the dog who dumped it, and the owner who neglected to pick it up can be fined. It’s a two-year trial project initiated by Béziers mayor Robert Ménard, who seems to have had his shoes soiled one time too many (“I can’t take any more of this mess,” he told French radio). The story ran in the newspaper the Guardian, where we hope it was a scoop.
Food for Thought
WGS, Part 1: A great Dx tool that may depend on the skill of the user
A recent study in JAMA reminds us that when it comes to using complex tools, the skill and knowledge of the user often matters a great deal. The paper presents the results of the Genomic Medicine in Ill Infants and Newborns (GEMINI) study: 400 hospitalized infants underwent WGS and simultaneously were tested with targeted gene panels, to see which method was more successful at diagnosing them. WGS was the clear winner (49% diagnosis rate versus 27% for the targeted panel), though results typically took a couple of days longer to arrive. About a third of the time, the diagnosis was a disease that the baby’s physicians hadn’t considered, and in some cases, getting a diagnosis for the baby helped to diagnose a parent, too.
The authors noted concern over a critically important wrinkle, however. In more than 80 cases, the two labs found the same genetic variant, but one lab’s geneticist reported that the variant caused disease, while the other lab’s geneticist labeled it “a variant of unknown significance.” The upshot: A genomic test of any kind is only as good as the geneticist interpreting the results.
WGS, Part 2: As a newborn screening tool, it’s fraught with ethical issues
Several broad programs to perform whole-genome sequencing (WGS) on heel-prick blood cards as a screening test for well newborns are either underway or about to get started. In the US, there’s BabySeq and Newborn Screening in Genomic Medicine and Public Health; the UK plans to begin their 100,000-baby Newborn Genomes Programme soon. A cautionary editorial in Lancet looks at these with a gimlet eye and weighs the pros and cons. (It also contains an excellent bibliography for those who wish to explore further.)
On the positive side, 4,000 single-gene mutations are known to cause disease, and early identification can provide prognosis and treatment (for an eloquent parent’s opinion, see today’s Washington Post). On the negative: Do we really want everyone’s complete genome on file somewhere? Should parents be able to provide consent that may affect their children’s entire adult life in ways no one can predict? False positives are far more common in NGS results than in more limited tests that are fine-tuned to identify the most common diseases - what about them? The authors’ conclusion: “The risk-benefit balance remains uncertain.” Proceed with caution.
If you build a training program for non-lab testers, will they come?
The CDC recently launched a training program for anyone - professional or volunteer - doing point-of-care testing in non-laboratory settings. It’s called OneLab TEST (Timely Education and Support of Testers), and is part of the agency’s larger OneLab initiative, started in 2021 and “developed to bridge, train, and sustain a capacity-building community among laboratory professionals and testers to collectively support rapid, large-scale responses to public health emergencies.”
Commentary: This sounds like a great idea in terms of epidemic / pandemic preparedness and acknowledging the importance of diagnostics, but it’s hard to imagine that people are lining up for this sort of training these days.
Is anyone testing outside of labs anymore? Pharmacies, probably. Workplaces, rarely if ever. Schools, maybe after holidays and for symptomatic students and staff. Long-term-care facilities, maybe (at least we very much hope so).
It’s good to see that CDC has been doing focused OneLab outreach to labs in rural and low-resource areas. We hope that they continue similar outreach to folks outside the laboratory community. Even if uptake is low now, the program has the potential to be hugely helpful in a future emergency.
Quick Hits
COVID susceptibility, the good news: People with a particular mutation of human leukocyte antigen (HLA) are more likely to be asymptomatic if they get infected with COVID, because they have memory T cells that are able to recognize a particular particle on the virus.
COVID susceptibility, the bad news: People with Type A blood are 20 - 50% more likely to contract the virus than people with Type O blood. However, they are no more likely to get a severe case of the disease.