NAME: Christine Ginocchio
POSITION: Senior medical director and chief, Division of Infectious Disease Diagnostics, Department of Pathology and Laboratory Medicine, North Shore-LIJ Health System Laboratories; and professor, Hofstra University North Shore-LIJ School of Medicine
North Shore-LIJ Laboratories is the core diagnostics facility for the North Shore-Long Island Jewish Health System, one of the largest hospital networks in the New York metropolitan area and in the US.
In 2009, the laboratories handled the majority of diagnostic testing for the swine flu outbreak that began when some 150 students at New York's St. Francis Preparatory School in Queens reported respiratory symptoms consistent with infection from H1N1 influenza.
That day marked a tipping point in North Shore-LIJ Laboratories' transition to molecular testing from rapid antigen testing to diagnose influenza and other respiratory viruses, and the laboratories have been at the forefront of testing and adopting new molecular diagnostic technologies ever since.
Christine Ginocchio, director of microbiology and molecular diagnostics at the core facility, was there on the weekend of the H1N1 outbreak and has played a central role in the laboratory's recent efforts to decentralize molecular testing. PCR Insider recently caught up with Ginocchio to discuss these topics and more. Following is an edited transcript of that interview:
Tell me about a problem that you have encountered in the clinical laboratory, and how developments in PCR-based molecular testing have solved that problem.
We're a very large core facility that serves 11 hospitals in our health system. It's been [more than] three years since the H1N1 [influenza outbreak in New York]. Back then … a lot of the molecular testing, other than something simple like a [Cepheid] GeneXpert, was really very complicated, and was not going to be done at the local site because of the complexity of the testing.
Almost all of our emergency departments were using rapid flu antigen testing, like everybody else was, [to test] patients. We were the epicenter of the New York City outbreak. Our hospital system … treated many of the students … from the preparatory school where the outbreak started. On Friday, April 24 — I remember the day quite well — we went from winding down flu season with about 150 flu tests per day … to doing 1,000 tests per day on Monday. We were overrun.
We [also helped] publish one of the first articles about the poor performance of the rapid antigen flu tests. We've had many discussions with the [US Food and Drug Administration] on this. A test gets … [in vitro diagnostic] clearance, and there is really no regulatory process, especially if it is a CLIA-waived test, that requires submission of any ongoing performance data for those assays. A lot of these rapid tests are CLIA-waived, low-complexity, and that's it. They never have to change, and they never actually lose their FDA clearance unless a major complaint is brought against them.
So one of the problems, of course, was that [H1N1] was a new viral variant. Many of the rapid tests had a poor sensitivity for detecting the virus, because it was a new strain. With influenza viruses, as with other RNA viruses, this is a major problem — they change very quickly. These genetic changes can not only affect rapid antigen tests but also molecular-based tests.
At that time we had just brought in the Luminex [xTag Respiratory Viral Panel]. It's a very complex test, and certainly would not be done in many local hospitals. It takes about eight to 10 hours to complete. But it was the first FDA-cleared test that had the ability to do high multiplexing and could sub-type the viruses, and potentially select out H1N1 from seasonal H1 and H3.
We had already validated this test; therefore we … immediately instituted it, and it really became a savior for us trying to deal with this large pandemic; to quickly identify … the scope of the pandemic in NY for the public health officials. We worked very closely with the [state and local] Departments of Health … deciding whether or not to close schools based on results. It was really an invaluable tool, and we've been using it ever since as the mainstay of our respiratory virus testing. We run the full panel, even the viruses that are not included in the FDA-approved version, and the test has provided a tremendous amount of epidemiology [and] outbreak information. We actually provide a weekly update to the entire health system on all the viruses that are circulating and their prevalence.
But in the meantime, tests have changed. Because of the complexity of [the Luminex] test, everything is centralized here to the core lab. At all of our hospitals, and especially our tertiary care centers, during the winter months, when we see lots of respiratory viral infections, patients have to be isolated because we don't know what viral infection they have. It becomes a logistical bed issue, and you don't want to cohort patients that may not have the same viral infections. A lot of infections can be dual or triple infections. We have to isolate patients, wait for our results to come back, which could take a day, and then decide what to do with the patient. Even though it provides really important information, the time delay has an impact in a very bad respiratory season.
We've now made the decision that you need rapid identification at the point of care: at the hospitals, for the patient in the emergency department and those patients who will be admitted to the hospital. So we are looking now toward decentralizing respiratory virus testing … [using] the Idaho Technology [FilmArray] platform [Ed. note — Idaho Technology is now called BioFire Diagnostics. See PCR Insider, 8/30/2012]. It's comprehensive – all of the viral pathogens are approved on their platform. You get the atypical pathogens also including Chlamydia pneumoniae, Mycoplasma pneumoniae, and even [Bordetella] pertussis, which clinically is a very different disease, but it's in the mix.
Now, running a [FilmArray] on a patient in the ED, in an hour and 15 minutes with literally two minutes of hands-on time, the local community hospitals and the tertiary care centers can provide a very comprehensive, accurate result very early on, so if the patient is being admitted, we know exactly where to put that patient, and who can [be cohorted] together. And you'll also know whether that patient leaving the ED should be put on oseltamivir [TamiFlu], or shouldn't be. We do not want to treat persons with oseltamivir when they don't have flu. There are a lot of other respiratory pathogens. They mimic each other. They can be very hard to differentiate clinically.
This is what you're going to see: whereas molecular testing was [once] centralized into sophisticated … core laboratories, testing is now going back out to the community, to the point of care, where a lot of it needs to be.
The FilmArray is a combination of multiplex PCR and high-resolution melt curve analysis, right?
It's a two-step PCR. First is in one highly multiplexed PCR [reaction]. Then [there is a] second-round nested PCR [with] multiple primers per target. Detection uses SYBR Green and PCR products are distinguished based on using endpoint melting curve data analysis.
Is the FilmArray platform truly point-of-care?
It's funny, there was just very interesting debate on the terminology of point of care at a recent meeting. We had a panel discussion, and somebody brought up a very good point: When you say point of care, you're thinking about a physician's office or something like that. But it really should be described as 'point of impact.' I thought it was a very clever term, because [the test] could be for an ED patient, but theoretically you don't need to do it in the ED. You could do it in the laboratory across the hall. Where are you going to have a rapid enough result so that you'll be able act upon it with that patient? In my mind, it doesn't necessarily mean … that I have to run up and do it next to the patient's bedside. I look at it much more broadly, where a lot of this testing used to be much more centralized, and we're bringing it back into the hospitals, where they have a rapid result for the patient. So point of care could be in a physician's office, it could be in an ED, or a stat lab. I think the term is getting a little bit clouded.
Also, something like a FilmArray is not CLIA-waived, so you're not going to be able to put it in most physicians' offices, unless they're under an article 28 of a hospital, and regulated by the state, CLIA, and sometimes [the College of American Pathologists].
Do you think getting these tests into a physician's office or a small clinic is a fruitful endeavor?
I have a little bit of concern about that. If they're simple enough, yes, it could easily be done. But … studies [and] surveys … on how good or accurate the results are … [have shown] that up to 50 percent of users in [physicians' offices] never read the package insert, or follow the exact instructions. They do not check expiration dates on reagents and materials. They let things sit longer than the allotted time to read a test – it might need to be read in 15 minutes, but they come back and hour and a half later. They do not collect the samples properly.
So if it's simple enough, it could be done there. But we have to be cautious and realize that problems arise with physician offices doing rapid antigen tests correctly, let alone doing something that, hands-on, may be comparable, but also has a lot of interpretive questions. Every positive doesn't mean that you act upon it.
It sounds like bringing molecular testing to the bedside could be an extra burden on physicians.
But it could be a big money maker … if they can bill for it. That's going to be the other limiting factor in [physician office laboratories] — that they need the codes to be able to bill for these in their offices. This is all changing with [Centers for Medicare and Medicaid Services], but say, for instance, you have a respiratory panel [with] 23 different targets — you could bill 23 CPT codes. Unfortunately, a lot of these targets don't have their own CPT codes, so you have to use this generic code. The reimbursement in New York State is approximately $49.50 per CPT code. So you're talking about getting back thousands of dollars for a single respiratory test. This is really not reasonable.
[Currently] in our reimbursement scheme the payors do limit the number of CPT codes reimbursed for these multiplex tests — usually they only pay for six per test, which is reasonable because that more than covers what the cost will be for us to run the test and even make a profit. CMS is looking to … lump multiplex tests into certain pockets. I don't know what the final outcome is going to be, but if it's one to three, or three to 10, then you're going to get a flat-rate fee for that. They don't want people with 23 targets billing 23 CPT codes at $50 a CPT code. That's just milking the system. The test may only cost you $150 … and if it’s an outpatient [that] doesn't have insurance, they are going to get a bill for $3,000 or $4,000 for a flu test, which is not a good thing.
If [physicians] can run a PCR test in their office, and some of the tests [cost as little] as $25 or $30 … if it's a single target and they get back $50, they're going to want to do it in their office. If the test is a multiplex and costs $150 and they can be reimbursed $300, then that is even a bigger incentive to do it in the office. It's a way of supplementing their income. In flu season … in a pediatrician's office … they could do 50 or 100 tests per day [and] that's a lot of profit. So that could be incentive to [run these tests] in their office.
There will be CLIA-waived molecular tests approved this year, and that makes it a whole different ballgame. Then you could truly have POC in that you could put it in a mobile van, a clinic, a physician's office.
Are you referring to any specific tests?
It came up in our discussions at the meeting, but I don't have specifics on any test. There are tests that have been submitted to the FDA as point-of-care molecular that will probably be approved this year.
Back to your core lab, is testing for respiratory infections the largest portion of what you do?
For molecular, no. We do a decent volume — probably 30,000 or 40,000 tests per year — for respiratory viruses, [using] a number of different [methods]. But we have much bigger volumes for molecular tests: chlamydia, gonorrhea, [human papillomavirus], [Clostridium] difficile – some of those are very high-volume tests for us, as well.
Do you find that the need for point of care testing is as strong for some of those diseases and pathogens as it is for respiratory pathogens?
No. HPV is not a stat test. How we decide whether or not a test should be centralized or moved on site is whether or not somebody is going to act on that result when it is available. If it is a test that is not critical for the immediate management of the patient then it does not have to be done at the point of care. … For example, HPV, you've probably had it for 10 years, and it's not going to make a difference whether I tell you today or tomorrow that you're HPV positive. It really varies. As long as you have a decent turnaround time for your testing, then it can be very easily centralized.
Is your lab eyeing any other new molecular testing technologies coming down the pipeline?
We've probably worked with every platform available. I do a lot of clinical trials and assay design and development, some with companies. We're doing a lot of work developing assays on the BD Max system. We've also been working on the Curetis [Unyvero] system … and hopefully Biocartis within the next year or so.
One of the new platforms that I really liked is a system [from Québec City, Canada-based startup] GenePOC. That's very clever. It's very small, the size of a FilmArray, but the computer is built into the cycler. The problem with a lot of these platforms is you need a separate computer, and then you need [some other piece of equipment], and have to stack multiple units — that's a nightmare. One of the limitations of the FilmArray is that it's one at a time, and for every test you have to have a laptop. They're not conducive to higher-volume testing. The GenePOC can run eight cartridges at a time and each cartridge can multiplex up to 12 analytes, and it only takes an hour. It's minimal [hands-on time]. You put a sample in the loading bay, you snap the cassette closed, load it in the machine, and then turn it on.
Is Cepheid's GeneXpert platform on your radar? That seems to be one of the first systems to have some of these features you are talking about.
We don't use it here in the core lab, because our volume doesn’t warrant using this type of cartridge system, especially for the very high-volume tests. But we do have some GeneXperts in some of our rapid response laboratories. They do quite well there. Some of the tests that we looked at that were lower volume and might be conducive to running a GeneXpert here at the core — we found their prices a little on the high side. We can do it much cheaper with the current assays that we're using. But we do have [GeneXpert] C. diff and MRSA at a couple of sites. It's a fabulous technology, and I'm very supportive of it. They were the first ones out of the gate to do a really terrific job.
What is on your short wish list for improvements to molecular diagnostic technologies or platforms?
I look at both the platforms and the testing menus. In the testing menu arena, we think of things as syndromic-based. If you come in with a respiratory infection, we're going to run X, Y, and Z tests, in multiplex. If you come in with gastroenteritis, we want a panel that's going to contain the bacterial, viral, and most common parasitic pathogens. That's how we want to look at approaching testing. We don't want to look at it as a single test for this or that.
What's going to really change a lot of what we do is [comprehensive panels]; for instance, in the GI family. If the panels are comprehensive enough, they'll replace what we're doing, not be an add-on. A lot of molecular testing now is an add-on test. If you had a very comprehensive stool panel — and some of the new panels coming out are not truly comprehensive, so you still have to do supplemental testing — you could replace [current testing]. Rather than having a stool culture in the lab for a few days, you'd do a one-hour PCR and be done with it. That's much more cost-effective; and you'll get much better results because it will be much more sensitive.
What we don't have are FDA-cleared panels for the meningitis/encephalitis groups. That's because they're extremely difficult and expensive clinical trials to do … because many of these [pathogens] are not very common.
But there's still a need because of the potential severity of some of these diseases?
Yes. We all have home-brewed tests for some of these. It would be great if we could have a good FDA-cleared test. The FDA is really conscious of that, and we're trying to come up with really creative ways to get these kinds of tests through the [regulatory process].
From a platform standpoint, 10 years from now PCR is not going to be where we want to be for doing this type of testing. There are some new technologies [from] biosensor companies … that may have the ability to have the same kind of sensitivity without the PCR up front. Next-gen sequencing holds a lot of promise, but the bioinformatics on that is nowhere near ready for clinical diagnostics. We were actually looking at some of the systems, but we need five people just to do the data. But that's the direction we're moving toward, we just need informatics and very well-vetted databases. Running the tests on these platforms is very simple.
One other thing that is revolutionizing infectious diseases is … mass spec, MALDI-TOF. That's completely changing how we do certain things in microbiology, which is a little bit behind the times compared to virology, where it's all molecular. I have worked with collaborators … to develop molecular panels to identify bacterial pathogens for sepsis, from positive blood cultures. We spent all this time and effort, and the assays were great. But now I can take a positive blood culture, spot it on to my MALDI-TOF slide, and in 15 minutes have a correct and comprehensive answer.
Nanosphere this year came out with a very nice gram-positive panel for their Verigene system, directly from blood culture, and maybe they'll come out with a gram-negative panel next year. I think they'll have an excellent place for use in smaller hospitals … but many very large hospitals, I think, will be moving toward MALDI-TOF to do that kind of identification.