Charna Albert
September 2025—Avian influenza, self-collection, and diagnostic stewardship in the microbiology laboratory are three topics of many that can be explored at the Association for Molecular Pathology meeting in Boston this November.
Andrew Pekosz, PhD, heads a research laboratory at Johns Hopkins University that studies the replication and disease potential of emerging respiratory viruses. In “Emerging public health threats: avian influenza A(H5N1) and the return of vaccine-preventable diseases,” Dr. Pekosz, professor in the Department of Molecular Microbiology and Immunology, will share his findings about highly pathogenic avian influenza and encourage vigilance as flu season approaches.
“The earlier we can find an H5-infected person, the better off we’ll be,” Dr. Pekosz says. Clinical laboratories should consider testing a subset of their flu A samples for H5, he says, although an H5N1 outbreak is far from likely. “But this is one of those preparedness things.”
Dr. Pekosz is studying whether the neuraminidase protein—the “N1” in H5N1—generates some cross-reactive immunity with avian influenza, given there is also an N1 in one of the human seasonal influenza virus strains, H1N1. “We’re trying to push the neuraminidase protein as a potential future flu vaccine target,” he says. “Right now, the vaccines are based on the other protein, the H part of the equation. And generating antibodies to H5 has proved to be challenging. But because you have some preexisting immunity to the N1 protein, maybe boosting those responses and making them more broadly protective is a faster, more efficient way to get some level of population immunity,” he says.
Virologists have known for some time that the H5 virus doesn’t replicate well at the temperature of the human upper respiratory tract, which is about 33°C. At 39°C—the temperature of its avian and bovine hosts—the virus replicates and kills cells quickly, Dr. Pekosz says. “We think one of the barriers that might be protecting people is these H5 viruses haven’t learned to replicate at lower temperatures,” he says. A newer finding is that the virus infects human cells quite well at 39°C. “There’s an assumption that the virus has to replicate or change some of the proteins that it needs to interact with to effectively infect humans, but we’ve shown that it can infect at 39 degrees quite efficiently without any mutations,” he says.
Other points of interest for Dr. Pekosz: how well the virus can kill human cells, how quickly it can spread from cell to cell, and what types of cells it can infect. Historically, he explains, when avian influenzas have infected humans, they’ve infected the ciliated cells in the respiratory tract; human influenzas typically infect the ciliated and nonciliated cells. “We see that in our cultures now,” he says. “We don’t know the true implications of it, but it’s something we’re characterizing. If we do see a change, that might mean the virus is moving closer to a human pathogen.” Some human isolates of H5 viruses that have been studied have one or two mutations that suggest it might be evolving in that direction, he says. “But those mutations haven’t fixed yet while the virus is circulating in cows and birds.”
Dr. Pekosz would like to see greater investment in phenotypic prediction from genetic data. “Viruses are malleable,” he says. The scientific community might understand one or two genetic changes that could allow for human infection, “but what we need to understand is what are the different pathways that can get there,” he says. “The more we know about that, the more we can see the early signs that a virus is changing in a dangerous way.”
The H5N1 variant B3.13 has been circulating in dairy cattle since early 2024. When the outbreak first occurred, Dr. Pekosz says, virologists weren’t aware of how extensively cattle herds are moved throughout the country. Once they understood that, “it was clear, when you follow the mutations in the virus and where the cows were moving, that the entire nationwide outbreak was driven by one jump of one virus from birds to cows.” For a time that was reassuring, he says. “Because we only saw one jump. We didn’t see other jumps. So we thought, maybe this is one of those lightning strike moments that may never happen again.” But then, another variant—the D1.1—made its way into cattle. “This tells us there’s some sort of hole in the biosecurity at these farms,” he says. “The big question, though, is how did that exposure happen? And there really is no good discussion yet, in terms of understanding how those jumps went.”
Notably, the D1.1 variant has caused two human fatalities. “It seems to be more dangerous to humans,” he says, “so we think that’s the genotype we want to follow the most, based on the limited data we have now.”
Yet the current H5N1 outbreak in dairy cattle has not been as disastrous as virologists might have feared, he says. It has long been assumed that an avian influenza outbreak in mammals would pave the way for the virus to become a human pathogen. But the current virus, he says, is infecting bovine mammary glands, an environment that to the virus is less like a mammal and more like an avian host. “It’s still a high temperature,” he says. “It has some of the same attachment factors that the virus uses, so it hasn’t had to change its receptors.” Transmission has been driven largely by infected material moving from one cow to the next via milking equipment, he says. To truly adapt to humans, the virus would have to become a mammalian respiratory pathogen. “We’re fortunate that’s not how it’s transmitting in cows.”
That’s not to understate the real danger of H5N1 spreading in mammals, he notes. If it next migrates to swine, for example, a reassortment event could occur, given that swine carry many different influenzas. “While there’s not many flu A viruses in cows, there’s a lot of flu A in swine,” he says. “So if H5 did jump to swine, the chances of a reassortment event could also increase tremendously.”
Given the ongoing risk, he hopes that national diagnostic laboratories continue to invest in assays for H5 subtyping. “Once you have these larger diagnostic companies understanding the focus and investing in those tests, that’s when you can make the biggest impact in terms of improving surveillance for these novel types of influenza.”
Those efforts—and added surveillance from clinical laboratories—will be critical, he says, given the ongoing budget cuts at the federal level. “It’s unclear who’s going to be able to step in and do some of these extra surveillance efforts,” he says. “We used to rely on the CDC. I don’t know how much that can happen in the future.”
In the AMP session on self-collection, Dina N. Greene, PhD, D(ABCC), and Meghan Starolis, PhD, will discuss the opportunities and challenges as self-collection becomes more routine.
“This type of testing is needed to improve health equity,” says Dr. Starolis, senior science director at Quest Diagnostics. “That is clear in the research, in the data. And the challenges that arise are surmountable, but we need to acknowledge them and work together to figure out how we bridge those challenges.”
People of means, too, need better access to screening for sexually transmitted infections, says Dr. Greene, clinical associate professor, University of Washington. “Even beyond health equity, this is needed for the sexual health of all people,” she says.
Quest last year began to offer patients experiencing vaginitis and other symptoms of infection in the genital tract the option to self-collect a swab specimen. “We allow self-collection for vaginal specimens at over 2,000 patient service centers using the Aptima Multitest Swab,” Dr. Starolis says. Uptake has steadily increased, with more than 500 patients per week now taking advantage of the option. It’s also reducing burden on clinical staff, she says, especially in high demand or urgent situations.
Quest may in the future accommodate extragenital sites like the throat and rectum, she notes. “It is on our radar. We’re not quite there yet, but it’s something we’ve had customer demand for.”
In her talk, Dr. Starolis will review the regulatory categories for self-collection: observed versus unobserved and at-home versus in a health care setting. “All of these nuances are spelled out in the intended use statement. If you’re not familiar with those, it can be helpful to partner with someone who has regulatory experience,” Dr. Starolis says.
The FDA in May of this year approved the Teal Wand (Teal Health), the first at-home self-collection device for HPV. “We know there is a high percentage of patients who are not getting screened for HPV,” Dr. Starolis says. “The ability to combine HPV primary screening with at-home collection has the potential to significantly increase the number of patients who are being screened.” The challenge, she says, is linking patients who need a Pap or colposcopy with appropriate follow-up care.
Another challenge, Dr. Greene says, is making sure patients who use the at-home option stay within the appropriate screening intervals. To avoid detecting transient infections that might cause emotional distress, “it is very, very important that people are not overscreened,” she says.
Other opportunities and challenges will come with approving self-collection for a greater number of sample types.
“Certain specimen types are well suited to self-collection—urine, vaginal swabs, nasal swabs—and that’s what we’re offering right now,” Dr. Starolis says. “However, as we look to the future, we don’t want to limit ourselves to types of tests that can be offered on these particular sample types.” In addition to cervical specimens, for example, new self-sampling devices now allow for blood collection. “It’s likely in the future that we’re going to see an expansion of the types of tests we can offer on self-collected samples.”
Dr. Greene will discuss in her talk the optimal specimen types for gender minorities. For cisgender women, she says, vaginal swabs detect discharge-causing STIs in the vaginal tract more sensitively than a urine specimen. But for transgender men on testosterone, whose vaginal lining changes significantly with hormone therapy, the optimal specimen type isn’t yet known. “Since the vaginal wall thickens, cellular shedding may be affected,” which may influence the sensitivity of vaginal swabs relative to urine, she says. “We just don’t know.”
Though some specimen types may be more suited to certain patients, she adds, they shouldn’t be insisted upon, particularly for patients with a history of trauma. “Even though we can say this is a better type, people who do not want to collect a vaginal swab for whatever reason should be allowed to provide a urine specimen.”
Rarely does she see data indicating that self-collected samples are less accurate than provider-collected samples, though it is important to cater to patients who may have lower health literacy and provide instructions for collection in multiple languages. “If I have a set of clinician-collected samples and a set of self-collected samples from the same exact patients, I’m going to have a higher positivity rate in the self-collected samples almost every time,” she says. “We know our own anatomy.”
Most microbiology laboratories are taking steps to manage and optimize resources, especially when it comes to the more nuanced and expensive molecular tests. But many aren’t documenting and formalizing those efforts, says Rebekah Dumm, PhD, D(ABMM).
“They’re in it to solve the problem. They’re not necessarily in it for the academic productivity,” says Dr. Dumm, assistant professor of pathology and immunology at Washington University School of Medicine in St. Louis and a medical director of microbiology and molecular infectious diseases laboratories at Barnes-Jewish Hospital. As test options in the microbiology laboratory have expanded, the scientific literature on diagnostic stewardship has struggled to keep up, she says, “particularly with the practical approaches people are taking.” But documenting an intervention’s successes will support the need for institutional resources to continue and grow those efforts. “Otherwise, you’re constantly playing catch-up as different tests come up or workflows change,” she says. “It’s important to document, document, document. And this isn’t something all labs are necessarily incentivized or resourced to do.”
In “Optimizing diagnostic stewardship: clinician and lab director perspectives on effective test utilization,” Dr. Dumm will review what’s been done in diagnostic stewardship “with a focus on molecular and multiplex testing,” she says. “The idea is to balance clinical utility, patient population needs, and resource management.” She’ll also highlight opportunities for laboratories to publish their own data.
Two recent American Society for Microbiology laboratory practices subcommittee reports discuss diagnostic stewardship, she notes. One focuses on the operational and regulatory justification for the laboratory’s role in stewardship activities (Dumm R, et al. J Clin Microbiol. 2024;62[10]:e0096024); the other offers guidance in implementing a molecular diagnostics stewardship program (Valencia-Shelton F, et al. J Clin Microbiol. 2024;62[11]:e0094124). “These resources tackle this topic from a very practical standpoint.”
As a rule of thumb, high-volume testing is often a target for diagnostic stewardship. Optimizing the test for the correct population is a good place to start, Dr. Dumm says, citing respiratory testing as an example. “You have a lot of options there and you’re talking about the differences between inpatient, ED, and outpatient settings,” as well as point-of-care testing. At Washington University and Barnes-Jewish Hospital, a multiplex syndromic panel for pneumonia became the focus of stewardship when retrospective data generated by several interdisciplinary teams and examined as part of an antimicrobial stewardship effort found the panel was being used inappropriately. Stepwise interventions included adding interpretive comments for de-escalation of therapy and cutting down on repeat testing when it’s being ordered by default, she says, “rather than as a thoughtful choice to order it.”
How to make such interventions sustainable is an ongoing challenge. “We know from the published literature that single interventions are not particularly long-lasting,” she says. “There may be an immediate effect, but with time or shifts in ordering practices or any manner of other things, we see it bounce back up or vice versa.” Education alone isn’t enough, particularly as a one-time measure. “It has to be consistent, creative, and that’s resource intensive,” she says. A goal for the field, she says, is to find other ways to communicate the value of a test or redirect to another without adding unnecessary burden to clinical teams.
Most stewardship interventions focus on the preanalytic phase. An analytic phase intervention might ensure that only validated specimen types are run for a particular test, while a postanalytic phase intervention could involve appending comments that direct to additional testing or a specialty consult. “Those are some of the less appreciated areas of diagnostic stewardship,” Dr. Dumm says. But generally, “if we’re thinking about resources, or not including the potentially confusing result in the medical record, then we want to focus on not performing the test in the first place.”
When it comes to genomic sequencing-based approaches, the strategy for stewardship can involve more than a hard stop, she says. One option is a carveout population, or establishing policies that sequencing-based assays are not routinely sent out for certain populations, like outpatients, but are sent without need for approval for other, high-yield populations, like transplant patients. Or clinical decision support might give interpretive context or collect information for a follow-up conversation with a laboratorian. Some institutions require a formal second signoff from infectious disease or microbiology, “or it might go to a panel of specialists to decide if that test is needed and give guidance for interpretation of results,” she says.
Diagnostic stewardship is often thought of as a cost-saving tactic, but it’s low on the list for Dr. Dumm, comparatively. “I don’t usually start with cost, and even if I make my way around to it, it’s from a perspective of how we can optimize the use of our overall resources,” she says. When it comes to sequencing diagnostics for microbiology, for example, “correcting over- or underutilization and providing context for interpretation is key,” she says. Even though the tests may be signed out by an expert, “institutions may not always have someone in-house familiar with the technology who is actively interpreting those results.”
“There’s a desire to help interpret these complicated reports or guide toward high-yield scenarios for this specialized testing that is independent of cost,” she says.
Charna Albert is CAP TODAY senior editor.