NEW YORK – With the coronavirus pandemic showing no signs of abating, test makers have been put on the front lines of finding out who has the disease so that clinicians can care for them, contain spread of the SARS-CoV-2 virus, and protect public health.
Tracking, validating, and evaluating tests is increasingly important, as desperate governments try to quickly procure and scale up diagnostic testing under tremendous time pressures.
The Foundation for Innovative New Diagnostics, or FIND, is a global health non-profit that supports development and evaluation of diagnostic technologies. The organization works with manufacturers on everything from R&D to increasing access to testing.
Cassandra Kelly-Cirino is the director of emerging threats at FIND. Kelly-Cirino previously worked on emerging infectious diseases at the Canadian National Microbiology Laboratory, as well as at the Wadsworth Center at the New York State Department of Health, where she served as the deputy director of the Biodefense Laboratory. Most recently, she served as vice president of infectious diseases with DNA Genotek. Kelly-Cirino also holds a doctorate in immunology and infectious diseases and has developed vaccine and immunotherapeutic candidates for anthrax infections.
In an interview, Kelly-Cirino explained what she and her team at FIND are doing to track all of the commercial products for COVID-19 diagnosis, and to ultimately advise global stakeholders on procurement.
Below is an edited transcript of an interview with Kelly-Cirino conducted this week.
When did FIND start looking at the COVID-19 situation?
We started looking into it in early February. We had been having discussions before that, but that is when we started to really ramp up what our plans were going to be, and try to understand where our efforts are best placed.
We usually do research and development to assist diagnostic companies to go and make tests, because most of the areas that we're interested in are ones that companies won't necessarily enter into on their own.
But we have learned from both the Ebola and the Zika outbreaks that's not the case for these larger pandemic-type situations. Companies will rush to produce tests. We very quickly decided we would not be supporting any R&D, and we would really focus on the evaluation of diagnostics, and try to aggregate the pipeline and make the data available to countries that want to rapidly implement tests.
Have there been any trials comparing any of the tests that are coming out? Or, are we still in a stage of everything sort of flying out of the labs quickly at this point?
Pretty much everything is flying out of the labs quickly. The first test that was developed was a lab-based test that then got transferred into one manufacturer through the Tib MolBio-Roche partnership. Most people have taken that up as the gold standard, or the reference test.
It is a great test, but the problem is that it is a manual test. It still requires an extraction procedure, it still requires a box, so it requires a lot of infrastructure. As this outbreak starts to scale, laboratories are quickly overwhelmed by the amount of work that goes into those kinds of tests.
Just two and a half weeks ago we started focusing on supplementing available data for other manual tests, so countries can actually buy other things besides Tib MolBio and we can relieve some of the supply chain pressure on a single manufacturer.
But what we've now realized is that as the outbreak has escalated in Europe, even in the high-income countries, there is stock-out of reagents and the labs are just absolutely overwhelmed.
We need simpler tests that can be implemented in a more decentralized manner, which is really what we associate more with lower- and middle-income countries that need broader access to simpler tools. This has become a world-wide problem. Now, we are also trying to look at the availability of data for the automated systems, and then also for the rapid tests, like the 15-minute pregnancy test-style strip tests.
There has not been a large clinical trial. There is no time for a clinical trial, really. There is time for some laboratories to do their own evaluations of tests that they want to bring in, and use that data internally.
We are trying to do an independent evaluation of a lot of different diagnostics, and then put all that data out there publicly. It is also taking us some time, and every day that goes by is too long. And so, we are also trying to figure out other ways of accessing data. We are asking people that have done their own evaluations of diagnostics if they would share that data with us, so we can aggregate it. If we can build a dataset, then that would actually start to inform national public health response plans in a way that could give confidence, in terms of the performance of the diagnostics.
How many tests are you talking about, globally?
We opened up an expression of interest (EOI) for molecular diagnostics and we had more than 220 submissions. And that is just the ones that submitted, so there are well over 220 assays out there.
Countries are getting called, and labs are getting called by companies saying, 'We'll give you our test for free, just test with it.' But without knowing the actual performance, it is a risk to put those into place for patients.
We just closed our immunoassay EOI, so our rapid diagnostic tests, and we got close to 100 submissions, which is an enormous pipeline of tests. We could not evaluate all of those simultaneously, so we are prioritizing which ones will get first evaluations partly based on their global availability, their supply chain, and their manufacturing commitments, as well as being able to supply to the entire world, specifically low- and middle-income countries, not necessarily just to certain countries.
What is FIND's role in global commercial diagnostics development, generally?
We operate very independently from individual companies. We are going to evaluate diagnostics that companies have agreed and committed to being able to make globally available.
There are some discussions that some US-based companies will only manufacture and supply [COVID-19 tests] into the US. If that is the case, then we likely would not prioritize evaluating them, because the US should be able to manage that on their own.
We're really meant to be supporting the [LMICs] and the countries that have difficulty accessing reagents, platforms, and data on their own. That is where we are going to aggregate that data and do the independent evaluations to inform countries on what to procure. We are working with the World Health Organization to make sure that the data we aggregate and develop becomes available to them as they try to decide who to list on their emergency-use listings. We are also working to try to help inform national governments and The Global Fund on the different performances of tests so that they know which ones to prioritize for procurement.
That is our main goal and main role in terms of manufacturers, to supply those data.
FIND is tracking the pipeline of projects for COVID-19 diagnostics – how many people are working on curating this burgeoning list, and how are you managing to stay on top of it?
The pipeline tracker and the list we are curating, we are doing with an extremely small team. FIND as a whole isn't very big, and we have a small subset of our team that has been freed up to do work on COVID-19. It is quite a bit of information, but we're trying to standardize how we collect the information so it is easier for us then to group and present things in a reproducible way.
We interviewed an emerging infectious diseases researcher in Hong Kong, Leo Poon, who developed one of the first SARS-CoV-2 assays. He noted that during the H1N1 pandemic he got emails about a molecular diagnostic product being commercially available the day after the pathogen's sequence was decoded, and he emphasized it is basically impossible to make and validate a test that quickly. How can users even know whether a test is validated or if it is only in the proof-of-principle stage and maybe not even a good test?
That is the million dollar question. On our pipeline, we've asked for only commercially available products. That means they have gone to essentially a " design lock" stage, their test is not going through development, and they are not working at the bench making changes every day. It's settled, and they've got what we call "instructions for use." If you call them today and say, "I want to buy the kit," they would have a product number, and instructions, and they would be able to ship you that product.
There are lots of other companies that are still developing brand-new technologies and they are exciting. But the problem is that they are still in development, so they are not going to be ready for quite some time. The ability to manufacture with quality control, and at scale, doesn't happen overnight. That takes a company that has quite a bit of experience. While we are very supportive of all the research and development that is going on, we feel that it is too far out for us to focus on in terms of trying to address the global need today.
There was an announcement a few days ago that Spain will procure 640,000 15-minute tests, but the manufacturer was not disclosed. Are you aware of any commercially available rapid tests that are can be produced at that volume that are reliable enough for anyone to invest in such a large volume?
Well, those are two different questions: "Are they available at that volume," and then, "Are they reliable."
In our pipeline there is a list of immunoassays, and there are 34 manufacturers that make 15- to 30-minute rapid strip tests. The question is, what is their manufacturing capacity, and do we know whether or not they work?
Manufacturing capacities are something we are trying to assess now, but we have heard from several of the manufacturers in Asia that they have quite a substantial manufacturing capacity; 250,000 per day is possible, so an order for 640,000 is high. I am a little bit surprised, but it is not out of this realm. Also, we don't know how long it will take — just because they are ordering it, they may not get all that in one shipment tomorrow, it may be that they get 100,000 tests each week for the next six weeks.
The question is, which test are they ordering, and how do we know that it works? That is what concerns me a bit more. Right now, I would say there are almost no data to support whether or not these tests work, and how they work.
There are two different kinds of rapid tests. There is one that detects viral particle that is circulating in your nasal and lung secretions. Those could be used to tell very quickly whether or not you are actively infected. And then there are other ones that test the antibodies that are in your blood. If you've been exposed, you will mount an antibody response. But there are several days in there where you are infectious, but you won't have detectable antibodies. If you use an antibody test to screen patients, and you call them negative, but you are in that window, then you are actually potentially releasing people back out into the community who think they are negative, but they are positive and they would be transmitting.
It is not a failing of the test, it is a failing of the use of the test. We have to be very clear about which test are they using, and how sensitive they are — are they going to pick up everybody that is infectious or are we going to miss 20 percent? If they are only 80 percent sensitive, what does that mean in terms of your confidence level? Do you want to test patients twice over a three-day period, because maybe over time they become positive?
These are the questions with these tests, which we also wrangled with on the molecular side, but not to the same degree because the molecular side is quite a bit more sensitive, and much more specific.
These rapid tests are fabulous, they are exactly what is needed, but because of that, people are going to start implementing them without having the data on how well they actually work.
And then the risk is, people test negative either because their body hasn't yet mounted an immune response or because the test wasn't accurate enough?
Right. Missing actual positives and putting them back out into the community, or back out into areas of contact with people.
Is there any chance this could be referring to like a 15-minute lateral flow isothermal molecular test?
I would be quite surprised. We haven't seen any put forward. We've seen isothermal proposals put forward, but they still require machines. My guess is, if anyone is ordering 640,000 15-minute tests, they are ordering lateral flow [immunoassay] tests. Those require binding. And the challenge is, they are fast, but they are never as sensitive and specific as a molecular test. In my opinion, that is completely fine, we just have to understand how to use them appropriately so that we don't make bad decisions from a healthcare perspective.
What else is needed now in diagnostics for COVID-19?
I think we really need to figure out how to scale and use the RDTs as quickly and as widely as possible. I think that the automated nucleic acid amplification systems obviously are also critical for health systems to be able to respond. My only other question is about the other reagent suppliers, like the swab manufacturers who have to make more swabs so we can collect more specimens.
This is really an intricate web of suppliers, and we are starting to see unprecedented demand on their supply and on the raw reagents that go into testing. From a global perspective, there are questions around how many tests should be supported, so that we are not splitting up raw materials across too many different developers. But then, how do we do that in a rational way so that we ensure global supply? I don't have the answer. I just know that these are the questions that we are struggling with as we talk to developers who say, 'Well, I have enough raw reagents to make 20 tests, but then that is all. Although I have the capacity to make a million, I don't have the reagents to put into it.' It is just an unprecedented scale.
In the US, there has been a lot of conversation about why people can't get tested when they want to, and it is possible people don't understand restrictions on testing, or that if there are a limited number of tests, health practitioners have to ration them and decide who should get tested first. Does it seem like this is maybe a public health perspective that is just not common knowledge?
We've just never had to really face anything like this. These are the types of prioritization questions that get put to groups like Doctors Without Borders in the middle of a war zone.
We don't operate under those types of circumstances. Lack of access to care isn’t usually caused by an overwhelmed local or national health care system, or an overwhelmed global health system.
Now, with limited resources, and limited staff, and limited reagents, it is a big challenge to answer the question of who gets testing. I can't answer it, and I feel quite helpless because of it. All I can do is say, the more information I can give on the simplest-to-use tests, the better off we are in trying to make more testing available to people. That is really our goal at FIND, to try to make sure that we have enough information as possible to use as many tests as we can, to have a positive impact on the outbreak.