Multiplex respiratory virus panels, which use PCR technology to detect up to 20 or 30 pathogens in one swoop, may not be able to pinpoint the virus infecting every patient with respiratory symptoms. But they’re getting close.
For among the impressive achievements of the past decade has been the discovery of eight new respiratory viruses, says James B. Mahony, PhD, director of the Regional Virology and Chlamydiology Laboratory and professor of pathology and molecular medicine at McMaster University in Hamilton, Ontario. “I think there are a few new pathogens out there yet, but we’ve just about found them all.”
It’s one among many reasons why multiplex respiratory virus panels, or RVPs, are beckoning more labs to make the leap into molecular testing. “Adapting to new technology is always slow, but labs are really starting to see the value in these multiplex assays,” Dr. Mahony says. In a very few years, he thinks multiplex RVPs will be the standard of care.
Only one large multiplex respiratory virus panel, the Luminex xTAG RVP, is approved by the Food and Drug Administration for clinical use. But in just the past few months, the FDA has approved smaller panels from GenProbe (Prodesse ProFlu+ and ProParaflu+) and Nanosphere (Verigene respiratory virus nucleic acid test, or RVNATSP). The FDA has granted emergency use authorizations for several others, including Cepheid’s Xpert Flu A, GenProbe’s Prodesse ProFlu-ST, an influenza virus detection and characterization panel from the CDC, and Simplexa’s influenza A H1N1 (2009) test from Focus Diagnostics. With many other multiplex RVPs already approved in Europe and Canada or in the regulatory pipeline in the U.S., a full-fledged competitive market appears poised to break out soon.
It’s happening with a major assist from the swine flu pandemic of 2009, which cast a spotlight on multiplex RVPs. “The pandemic was a great opportunity and showcase for multiplex technology, and particularly for Luminex,” says Gregory Storch, MD, professor of pediatrics, medicine, and molecular microbiology, and director of pediatric infectious diseases and director of laboratory medicine in pediatrics, Washington University School of Medicine. “That was because Luminex had the capability to subtype influenza A and show a pattern—positive A and negative H1 and H3—that indicated that a sample probably contained novel H1N1.”
Even many holdouts are conceding that sooner or later, they’ll be acquiring a multiplex RVP. Multiplex platforms could provide the final nudge for the largest percentage of laboratories that are at the stage of seriously thinking about offering molecular testing, says Richard L. Hodinka, PhD, director of the Clinical Virology Laboratory, Children’s Hospital of Philadelphia, and professor of pediatrics, University of Pennsylvania School of Medicine.
“These laboratories have a strong desire to bring it in,” he says, “but most lack the resources and infrastructure. Commercial multiplex platforms could really change what these laboratories can do.” In the pandemic’s wake, Dr. Hodinka heard anecdotal accounts of hospital administrators who earlier were reluctant to put up the resources needed to perform molecular testing and then all of a sudden gave the okay for a multiplex panel.
At the same time, with the increasingly broad range of pathogens they can swiftly and easily detect, and an ease of use that couldn’t have been imagined 10 years ago, multiplex RVPs are taking both research and clinical care in unexpected directions. As Dr. Mahony sums it up: “The more we use multiplex RVPs, the more we find we should be using them.”
When Christine C. Ginocchio, PhD, set up two Luminex xTAG RVP instruments for routine testing last spring at the North Shore-Long Island Jewish Health System Laboratories, the timing was providential. She didn’t know that the 13-hospital system was about to become the epicenter of the Queens, NY, 2009 influenza A H1N1 virus outbreak. She and colleagues had been using the Luminex xTAG RVP assay and instrument for clinical trials and research, and they wanted to use it in the clinical lab for a broader scope of coverage with higher sensitivity.
“We didn’t want to switch to a molecular test in the middle of the normal flu season, so we planned on training and converting everyone in June when it’s relatively quiet. But when swine flu hit us, we had a week to get everything up and running,” says Dr. Ginocchio, senior director, Division of Infectious Disease Diagnostics, NSLIJHS Laboratories, and associate professor, Departments of Laboratory Medicine and Pathology and Department of Molecular Medicine, Hofstra University School of Medicine in collaboration with NSLIJHS.
During the two peaks of the pandemic, the instruments were in constant use. “You can run 96 at a time and we were setting up one run after another—sometimes more than 400 tests a day.” In the very beginning, that wasn’t fast enough because as many as 900 tests a day were requested during the height of the outbreak. Currently, the laboratory can guarantee its multiplex results in 24 hours.
In its current form, the Luminex RVP assay is more suitable to medium- to high-volume laboratories, she believes. “The xTAG RVP assay is not going to be put into a rural small community hospital. Smaller hospitals would need a test that is simpler to perform and does not require as much technical expertise, such as a Cepheid GeneXpert assay, Nanosphere Verigene RVNATSP, or a GenProbe Prodesse ProFlu+ assay,” Dr. Ginocchio says.
Dr. Hodinka’s laboratory at Children’s Hospital of Philadelphia made the move to routine multiplex RVPs about six years ago. Starting in the late 1980s when a lot of the respiratory virus antigen tests became available, “Our algorithm was to use a combination of solid-phase immunoassays and direct immunofluorescence antibody [DFA] tests for rapid detection of certain respiratory viruses, and if these assays were negative, specimens would then be processed for virus isolation in cell culture.” This type of algorithm involves considerable time, labor, and resources, he adds, and is not fully sensitive. But since the early 2000s, the laboratory has done no cultures and no DFAs, and uses rapid immunoassays in a limited way only for RSV. “Our current algorithm involves essentially just the panel of PCR assays,” Dr. Hodinka says.
His laboratory is one of the rare ones that have developed their own multiplex panels. It had been doing panel testing since long before any manufacturer had commercial multiplex assays on the market, he says. “As a large tertiary academic care facility linked with the University of Pennsylvania, Children’s supported the system because it wanted the laboratory to be state of the art and provide the highest testing quality and accuracy for patient care and management.” Their current PCR-based assays accomplish this, he says. They provide rapid and sensitive results that are used for outbreak ID and infection control; to help limit unnecessary antibiotics, laboratory tests, and hospital procedures; and to assist in the care of high-risk patients and guide antiviral treatment decisions.
Dr. Hodinka refers to the panel as a “multi-analyte system,” because on one given 96-well plate the laboratory will run 10 separate respiratory virus real-time PCR assays, though they’re not multiplexed in the same tube. Run in multiple small batches throughout the day, the panels are completed quickly, with most results being reported within the same shift and 100 percent turned over in 24 hours if no additional testing is required.
Luminex, by contrast, has efficiently figured out how to routinely multiplex large numbers of analytes. However, the technology is more the conventional endpoint type of PCR rather than real-time PCR, and the wide open platform of the Luminex system does create the potential for contamination, Dr. Hodinka points out. In addition, when using Luminex it is difficult to complete a panel in a single eight-hour shift. “Most laboratories will get the test to a certain point, then either hand it off to the next shift or wait until the next day to finish,” though Luminex is working on a method of reducing testing time and producing faster results.
John Brunstein, PhD, was an early evaluator of a different multiplex RVP, the Qiagen ResPlex I and II assays on a Luminex platform, and now an iteration of ResPlex II on the Osmetech XT8 platform. “We’ve been using it for about three years,” says Dr. Brunstein, clinical laboratory scientist at the Children’s and Women’s Health Centre of BC in Vancouver and clinical assistant professor, Department of Pathology, University of British Columbia, Vancouver. “It uses a method called Templex PCR, which is an approach to doing highly multiplexed PCR that avoids some of the kinetic and thermodynamic problems around high multiplicity.” It involves little hands-on time and doesn’t require a real-time instrument—“just a plain old-fashioned thermocycler.”
The XT8 is amenable to a continuous flow. “There were bottlenecks at the thermocycling stage every three hours, so we got around that by buying a bank of three thermocyclers. So once an hour we can always start a full load of specimens,” Dr. Brunstein says.
His hospital doesn’t use molecular testing as a first-line test for economic reasons. “We still run DFA, which we can turn around in less than two hours with relatively good sensitivity for A, B, and parainfluenza 1, 2, 3. It’s very cheap and it’s fast. Then we take the negatives and run them on molecular assays.”
Certain immunocompromised patients or transplant patients go straight to the front of the line. “If we had the money, we’d love to take every sample that comes in and put it straight onto molecular.” Molecular turnaround time continues to drop, he says: “We can do the panel in 4 ½ hours without breaking a sweat.” But it has to come down in price to be practical, Dr. Brunstein believes.
One seemingly minor feature of the XT8 is handy in a clinical setting. “It works via a touchscreen interface. The techs normally have to use a computer keyboard as part of the process, so they have to take their gloves on and off because it’s very hard to clean the keyboard. The touchscreen we can wipe down with an ethanol wipe, so they don’t have to take their gloves off. It makes for a much smoother workflow.”
Osmetech’s “black box” nature can be a drawback, he says. “I can’t build my own assays with Osmetech. If there is no assay for what I want, I tell everybody to leave us alone and we’ll wander off in a corner for a month or two or three and develop a custom multiplexed assay using our two Luminexes. We have U.S. patents on some of those assays—which have been easy to develop on a Luminex platform.”
Contamination is a potential problem with both the XT8 and Luminex panel, but Dr. Brunstein says it’s avoidable. “Yes, you do open tubes, but as long you’re doing that opening in a different facility from where you’re setting up your PCRs, then amplicon contamination is not a problem.” His laboratory has separate, sealed, pressure-controlled areas for sample preparation, PCR setup, and hybridization mixing. “We’ve been running them day in and day out since 2006, and contamination is not an issue.”
The Cleveland Clinic is taking a more cautious approach to multiplex RVPs. One of the concerns is overuse of tests, says Belinda Yen-Lieberman, PhD, the clinic’s director of clinical virology, serology, and cellular immunology. Although the laboratory has two Luminexes for different research studies, “We thought that using RVP for the general patient population was overkill,” says Dr. Yen-Lieberman, professor of pathology at the Cleveland Clinic Lerner College of Medicine.
Three years ago, with the price of the Luminex RVP panel running at about $395, the pediatricians and oncologists at the Cleveland Clinic agreed to stick with the Prodesse ProFlu+, a three-virus panel (influenza A and B and RSV). The ProFlu+ test, which employs real-time PCR, is user-friendly and, together with automated nucleic acid extraction (EasyMag) and amplification/detection (ABI-7500), with less than one hour of hands-on time, is more than adequate for outpatient and inpatient testing, she says.
However, “we are now ready to move on to a multiplex RVP for immunocompromised and hospitalized patients, maybe next year,” she says, possibly Qiagen’s ResPlex II or the EraGen Multicode-PLx system, depending on the regulatory status of these instruments with the FDA. The EraGen is in transition to a new platform and not available now, but “the chemistry of the EraGen is very clever and it is probe-free,” she says. The detection system, which employs a BeadXPress instrument made by Illumina, has no washing and centrifugation steps. “The Illumina uses glass-bead VeraCode technology. It is a compact instrument, and it’s definitely easier to use than the Luminex,” Dr. Yen-Lieberman says.
For different reasons, the Yale University School of Medicine virology lab is taking a wait-and-see stance on the commercially available multiplex RVPs. Marie Louise Landry, MD, professor and vice chair of the Department of Laboratory Medicine and director of clinical virology, likes the ability to control which viruses are included in its laboratory-developed RVP panel, which uses real-time PCR with samples in 96-well plates. “We put into the panel the viruses that are circulating at that time of year and that are important. We do a respiratory virus screen DFA panel as our first line ‘rapid’ test on all respiratory samples, and usually confine respiratory virus PCR to inpatients or high-risk outpatients.”
Currently, human metapneumovirus is the most common virus they’re detecting. Rhinovirus is not picked up by direct fluorescent antibody and is common year round, so it’s in the PCR panel. “Adenovirus is in the panel because we know adenovirus DFA is only about 60 percent sensitive. Other viruses may be added as appropriate or on request,” Dr. Landry says. When swine H1N1 influenza was prevalent, it was included in the PCR panel, in addition to influenza types A and B. During the peak of the novel H1N1 pandemic, a second influenza-only PCR panel was run daily.
Although optimistic about the future of multiplex testing, Dr. Landry finds that the first generation has not been as sensitive as her own lab’s panel. “I think multiplex molecular testing is ideally the way to go.” But her laboratory compared the multiplex RVP on one such platform to its in-house PCR and DFA, and the multiplex RVP missed a number of DFA positives for RSV and human metapneumovirus and was not as sensitive as real-time PCR for other viruses. “So we didn’t go ahead with it. We were concerned we would have a drop in sensitivity.” There was a long turnaround time relative to real-time PCR, and cross-contamination was a concern as well, she says.
In addition, “when you get multiplex panel positives, you have no indication of viral load. In contrast, real-time PCR provides a cycle threshold, so you can assess viral load.” A very low level of virus can reflect a recent past infection, whereas a high viral load is more likely to reflect an acute infection. “This is particularly helpful in assessing clinical relevance when multiple pathogens are detected.”
For Dr. Landry to change to a multiplex RVP, the sensitivity would have to increase. “Then we’d have to see how we felt about dealing with the lack of quantitation.” Real-time PCR, she notes, in addition to being faster and semiquantitative, reduces cross-contamination because the tube isn’t opened. “The probe is put in with the primers. So once you put the reaction together and close the tube, you never open it again.”
Dr. Storch of Washington University runs one of the laboratories that are still on the fence but considering a move to the Luminex. “We still haven’t decided, but we are thinking we will set up either the Luminex assay or a different multiplex assay at some point in the near future,” he says. Expense is a concern. “If I just substituted it for the other respiratory work we’re doing, it would have the effect of increasing the cost a great deal.”
The laboratory now does a combination of fluorescent antibody staining for acute respiratory virus infection, then culture mainly using the R-Mix system, and more rapid tests for influenza and RSV. “I have observed that labs like ours that are heavily invested in fluorescent antibody staining tend to be a little slower to adopt multiplex RVPs than labs that aren’t doing DFA staining,” Dr. Storch says.
Switching to multiplex PCR could also mean a longer turnaround time. “We have to be a little careful, because it could actually degrade our turnaround time if we simply substituted it.” With staining, the TAT is usually four hours and sometimes as little as two, while multiplexes are six hours at least, and Luminex takes almost three days because the nucleic acid extraction has to be done first.
The Luminex panel is still not approved for many different specimen types, he notes. “It’s mainly only for aspirates and nasal pharyngeal swabs. We’d be very interested in using it on bronchial alveolar lavage specimens, but it’s not approved for that yet.“
Aside from the logistics and cost of using them, multiplex RVPs have stirred a certain amount of controversy because, among the respiratory viruses, only influenza A and B are actually treatable at this point. That has made the clinical value of multiplex RVP results subject to debate, but molecular testing experts suggest the issues are multifaceted.
Dr. Yen-Lieberman thinks 12 to 16 viruses will be available on the multiplex panel her lab acquires, but says the lab will end up using just eight of those—“almost like the old traditional viral culture panel we used to offer.” When she talks with clinicians, she finds they don’t think it’s clinically relevant to test for parainfluenza 1, 2, 3, 4 because there are no drugs with which to treat. Typing for coronavirus is not needed either, she says, because “we don’t do epidemiology here, strictly patient care.”
Dr. Storch is concerned that some of the information the multiplex panel provides may be difficult for physicians to interpret. “I’ve heard that when physicians get the report that, for example, rhinovirus is present in the specimen, they are not always certain what significance to attach to that. That’s one thing causing us to go a little bit slowly.”
However, in Dr. Ginocchio’s view, “Influenza A subtyping is one of the most critical portions of the multiplex panels,” because more than 99 percent of the seasonal influenza A H1s are resistant to oseltamivir. “For acutely sick patients, appropriate antiviral selection is important. If significant antiviral resistance develops for 2009 influenza A H1N1, the typing component is going to be even more important.” Moreover, with additional antivirals being developed and in clinical trials now for viruses other than influenza A and B, it will be important to have an accurate diagnosis, she says.
Preventing unnecessary antibiotic use, of course, is one of the chief goals of testing—particularly in pediatrics, where the standard now is not to give antibiotics for viral causes. Adults admitted to the hospital generally have more severe disease, and clinicians are more reluctant to discontinue antibiotics, Dr. Ginocchio notes. Dr. Landry has seen the same. “Over a year ago I went to the ICUs and talked to physicians in pulmonary and critical care to tell them we planned to do the PCR panel. We already had data to show more rhinovirus would be detected, and possibly more metapneumovirus,” Dr. Landry says. “I asked, will this change what you do? and they said no—because a lot of these patients are on respirators, are in the ICU or have pneumonia, and may have underlying disease.”
“The sicker the patient, the less likely that the physician will have the confidence to stop antibiotics. We just don’t have the data yet. Studies need to be done to determine the right course of action.”
Nevertheless, multiplex assays are starting to open our eyes about causal agents for respiratory symptoms, Dr. Brunstein says. “Historically, we had testing for single pathogens, where you start going down a list, and when you get something likely then you stop. With a multiplex test, all the tests run simultaneously and the cost is the same for one result or 30.”
That produces much broader data that turn out to be very useful. In studies of thousands of patients between 2006 and 2009, “we found a lot of patients don’t have just one pathogen. There’s increasing evidence that certain bacteria or viruses tend to co-associate. For example, RSV B and strep pneumonia have a very close co-association. RSV B makes the patient a better receptor of bacteria, so if you detect RSV B, you really ought to take a look,” Dr. Brunstein says.
That doesn’t mean the molecular laboratory has the answer for everything. “When we get results and two or three things are positive, the physicians say, ‘What do we do?’ And we can’t always tell them,” he says. “Do you recommend antibiotics or should you wait a couple of days to see if it’s a virus that abates on its own? It’s something we have to collect data on for 10 or 20 years to understand.”
Dr. Landry thinks the RVPs could have the biggest impact on outpatients, relatively healthy people who come to clinicians with respiratory symptoms. “If there were a sensitive multiplex assay that could rapidly and economically identify viral infections at the point of care and prevent unnecessary antibiotic usage, it would have the greatest impact.... But we’re a ways away from that.”
“I think that, like it or not, more labs will be using multiplex RVPs,” says Dr. Brunstein. “If someone comes in sick, and I get a negative test, I’m really worried I missed something important. Did I get a false-negative? Is this really treatable? Is it transmissible? Is it a new novel pathogen? If somebody comes in and I pick up a coronavirus OC43, I’m pretty certain I know why they’re sick and it’s probably self-resolving and doesn’t need quarantining or cohorting. So even though I can’t treat it, knowing what it is puts us to rest as to what it isn’t.”
And as more data start to come in, he says, “Let’s say we suddenly found our coronavirus occurring in 30 percent of people who were sick. Then you might have enough incentive to say: Could we develop a therapy for this virus? If we didn’t know, then we wouldn’t put the effort into developing therapies. That’s one reason we don’t have effective antiviral treatments. So this is going to affect our paradigm of ‘one presentation equals one infectious agent.’”
Dr. Mahony, who has been working with a Luminex panel since its inception about five years ago, says refusing to find out whether a patient has coronavirus because there’s no treatment is thinking too narrowly. Just knowing an infection is viral not only prevents unnecessary antibiotic use but also can relieve anxiety in a mother with an otherwise healthy young child, he points out, because if it’s coronavirus or rhinovirus, she can be reassured it’s a mild self-limiting illness and the child will recover in about seven days. On the treatment front, “Respirologists are keen to know whether a patient with an exacerbation of asthma has a viral infection to help them decide whether to treat the patient with an inhaled beta-2 agonist, a glucocorticoid, or alternative medications.”
Surveillance is another good reason to have multiplex RVP results that are as informative as possible, he emphasizes. “To know what’s circulating is very important for public health authorities.” When Dr. Mahony’s laboratory first started working with the RVP in January 2006, “we found we had 10 different respiratory viruses circulating in the community in that one month. Normally we see three or four; that was unheard of to have 10 at once. It opens the eyes of our clinicians in the various services of the hospital.”
At McMaster University, there was an outbreak recently in a residence of a new coronavirus called HKU1. “We had 10 isolated cases in one geographic area, and there was a similar outbreak of HKU1 at a hospital in Toronto. We would never have caught these outbreaks without the multiplex RVP,” he says.
Multiplex testing is evolving, Dr. Landry says. “When people set up these tests, they try to design primers and probes that target conserved regions of genomes, rather than areas that are variable between different strains, to maximize sensitivity. But they’re basing their design on a limited dataset of sequenced viruses that have been submitted to GenBank,” the National Institutes of Health genetic sequence database maintained by the U.S. National Center for Biology Information.
“As people do more testing,” Dr. Landry says, “they often discover a given virus is more variable than they thought, or new strains appear. They did the best job they could at the time, but now they need to redesign the primers and probes to improve them. But you don’t know until you do it.”
It’s exciting, she says, that multiplex RVPs make it possible for viral infections to be diagnosed so quickly. “There are many viruses that are very difficult to grow in culture or don’t grow at all. When people started using molecular testing, they began to detect significantly more rhinovirus lower respiratory tract disease in hospitalized patients. Then when they sequenced the rhinovirus positives, a whole new group, rhinovirus group C, was discovered. Without molecular testing we wouldn’t know this group existed, since rhinovirus group C viruses have not yet been grown in culture.”
Important discoveries like these, by the way, can often be made with multiplex RVPs at no additional cost. Although the Luminex panel has 20 analytes, only 12 are now approved by the FDA, Dr. Storch points out. But the multiplex costs the same to perform whether one test is ordered or 20. While the instrument is doing all the testing, the results for non-FDA approved tests like coronavirus, enterovirus, and avian flu H5 subtype are often suppressed when clinical results are reported, but they are useful for research purposes.
As far as reimbursement is concerned, Dr. Storch says, “I think different hospitals are using different CPT codes. Some are using a single CPT code for each analyte and multiplying by the number of analytes included in the test so they get up to a very large charge. But I’ve also heard concerns that some third-party payers are not paying enough.” Some U.S. sites, Dr. Brunstein notes, have chosen to print their laboratory requisition forms with a single box to check that says “Respiratory Panel.”
In Canada, though laboratory budget squeezes are often cited, multiplex RVPs have been found to do more than recoup their costs, Dr. Mahony says. He wrote an article, published in the Journal of Clinical Microbiology last September, reporting that “we would save our hospital three quarters of a million dollars every year if we tested every patient with this test. We would prevent a lot of nosocomial infections, discharge patients faster, manage infections in the hospital more effectively, and eliminate unnecessary workups for bacterial infections.”
Coming in the midst of the subprime mortgage meltdown and the downturn in the world economy, that finding hasn’t made as big a splash as he would like. However, Dr. Mahony notes that Luminex has just released a smaller version of its instrument, called Magpix, which uses bead technology and will cost about $25,000 Canadian as opposed to the larger instrument, which is about $65,000 Canadian. This is going to help small labs make the leap to multiplex RVPs, he predicts.
“There are still 15 to 20 percent of patients out there with viral infections where we haven’t identified the agent yet. So these assays are moving targets, but the companies have done a remarkably good job of updating their assays,” Dr. Mahony says. Luminex and Qiagen, for example, are working on pandemic H1N1-specific assays and plan to add them to their panels.
It’s quite a shift, Dr. Hodinka says. “For years, you would often hear from commercial companies about how they planned to expand their menu and provide a host of PCR assays and multiplex panels. But their menus have stayed unbelievably small given that PCR has been around for so many years now. Because of the technical demands and costs, manufacturers of PCR assays have focused primarily on the highest-volume pathogens. Now the technology is becoming much simpler to use and getting better and better, smaller and smaller, and quicker and quicker, and the test menus are growing.” The companies have even more exciting innovations in the works, and he predicts major interesting developments.
Anne Paxton is a writer in Seattle.