COVID-19 Briefing for 21st April 2020

Notes from a conversation held by Lara Marks with Stephen Baker, professor of molecular microbiology, Jeffrey Cheah Biomedical Centre, Cambridge University on Friday 17th April

Progress in rolling out the tests

We are up and running. We are receiving about 100 swabs a day.

The NHS has set up a sampling service which seems a little bit cumbersome at the moment in the way they log in people and they are trying to change it because that is a bit of a hurdle to getting more people in. They get swabs and the swabs get delivered here at about 5pm. We then process them the next morning which means we extract the nucleic acids from them and then amplify them in the afternoon. We report the results on their server which uses their software. So, we can upload the results directly to them which means they can then be verified by somebody affiliated to PHE.

At the moment we are getting about 3-4% positives. That is not a massive amount, but it is enough to make the testing worthwhile.

Challenges to improving turnaround time

Our turnaround time is now quite good. It is 24 hours from taking the swab to then having someone verify the results in the microbiology laboratory in the hospital.

The limitation to us moving faster is getting more people swabbed and getting the swabs quicker and then somebody to then check the results. Actually, we could receive the swabs in the morning and then process them in the afternoon. That would enable us to report back the results the same day. It is just a question of logistics for when they swab people and when we receive them.

We are sending the extracted nucleic acids the same day to Ian Goodfellow in virology and we get back genome sequences for the positive cases the next day. So we can get a positive result and a genome on healthcare workers within 48 hours.

Within our laboratory we can turnaround the test in 4 hours. If we get a swab and extract the nucleic acids we can get a result out of the machine within 4 hours.

The reason it takes longer is because we don’t receive the swab until 5pm and people don’t want to stay in the lab until midnight. So we have to process the swab the next day. If we were to get the swabs early in the morning, we would therefore be able to turn around the test much faster. It would allow us to process them straight away which takes about 2 hours and then set up the machine and have a result. If we could have the swabs by 10am we could have a result out the door by 2pm.

Getting more support to speed up the process

The key issue at present is working out the logistics and the timetable. That is hard work. But we have had a call from various people, so the cavalry is coming. For example, Cancer Research UK, which is just across the courtyard from us, have offered to basically do what we are doing manually over there. So just by brute force with more people and space and raw materials Cancer Research UK reckon they can do about 500 samples a day. We are in negotiations with them at the moment about what the amount will be and then they will order the reagents and they will support us with the extractions. Basically, if that works, which it probably will because they have more staff and they have more space, then they can do the extractions and we will do the reporting. That will not only improve the process but be less of a demand on things here which means we can do other stuff.

Collaboration is essential to make things happen

We went around to look at Cancer Research UK’s lab yesterday and had a nice discussion with them. It is amazing how fast it was to get things going - if only things were like this all the time. Within 20 minutes we had looked round the lab and sealed the deal. This is what we need in science. In order to make things happen in science you need people who are just going to do it.

In terms of the blueprint it has just come out on BioArchive and loads of laboratories have already contacted me about it. I have shared it with people in laboratories in South-East Asia, India, Bangladesh and other parts of the UK.

We have established a great relationship with Public Health England. We are linked with the hospital computer system which is part of the PHE lab so they know what we are doing. They are fine about validating our results on their computer system. They are happy with our procedures. They get all of the results that we report and also the raw data too so they can check back and make sure they are happy with everything going on.

Finding a method to determine the immune response to COVID-19

Something we are talking about already is how we move from a diagnostic test to detect the virus to a serological test to know if a person has actually been exposed to the virus. What I would like to do is bleed the healthcare workers every time they come in to give a sample and link that back to whether we know they have had a known virus or not. If we can bleed the people, we know have had an infection with the virus, be that symptomatic or asymptomatically, then we can know their antibody response to it longitudinally. That would allow us to work out the proportion of people that have seroconverted - knowing the number of people who have been exposed to the virus. It would also help us to understand what that means and whether the difference in the degree of antibody response and whether you have had an asymptomatic or symptomatic event. That is something we still don’t know.

The issue is working out what the immune response is to the disease. Everybody is talking about the antibody tests, but I am not convinced there is anything out there that is particularly good. What we really need is the scientific work to correlate an antibody titre with a certain degree of protective immunity.

What we are talking about is working with Ian Goodfellow. We will work on the assays to look at the antibody responses. He is making a different virus phenotype that he will express the proteins on. That will allow us to see what antibody titre correlates with blocking activity against the virus, so we can tell whether the antibody is actually doing anything.

We have an antigen from people around here and we are trying to get our hands on some serum so we can get going on the issue next week. We have some negative serum already. So we now just need some positive serum and then we can start doing work to see if we can get an antibody against the virus in people who have had a known infection. Then we can get more samples and start screening them blindly and start demonstrating the curve to work out what the kind of natural antibody response is in the population.

Rapid turnaround test for ICU patients

Next week we will be able to talk more about what is happening in the Intensive Care Unit. Remarkably we are doing a rapid turnaround test using a PCR test based on a microarray card looking for 80 different pathogens in bronchial fluid taken from patients in ICU. The test looks for these pathogens in one go. Actually, it is working, as we can report a COVID-19 result and a diagnostic result for other pathogens in bronchial fluid for people on ventilators within 7 hours.

It is very impressive because at the moment it takes 72 hours. This is not because they can’t do it quicker, but it is because of the amount of time it takes - it is a bit cumbersome with the way we are currently doing it because it involves three labs. It involves getting the sample, doing the extraction, doing the COVID-19 PCR and then putting it on the TaqMan array card to detect the pathogens and then reporting back the result. The microarray-based test makes it possible to work out which patients have COVID-19 so they can be taken off the ward and treated accordingly without waiting 72 hours to get the result.

The test is based on the same process we developed with Vilas Navapurkar before the outbreak of COVID-19. What we are working on is reducing the turnaround time and adding the COVID PCR to it.

We are putting through 4 or 5 samples a day at the moment. The test can detect 80 different targets in each sample. This is done with the help of PCR which expands the number of infectious agents we can detect. While good at giving an early diagnostic result, the microarray based test is not yet a perfect process and still needs to be validated through other procedures. We still need to do microbiological culture on it. Nonetheless, the test is able to give you an early readout of the data of what is in the sample that you can follow up on.

Unfortunately it is difficult to quantify the direct benefits of the test. But the reality of it is that it allows the reallocation of resources and priority to other patients. This can be gauged from some observations with two patients, who when they arrived on ICU had not been confirmed to have COVID-19 or not. In one case the microarray test identified them to have the COVID-19. This result enabled the patient to be moved into an isolated treatment ward somewhere else, which freed up space for another patient to be accomodated. In another case a patient tested negative with the microarray test. As a result the team were able to move him to another room in a less higher dependency area which needed fewer nurses. This freed up the room for another patient. It also freed up the nurses who were originally allocated to him in case he developed COVID-19 symptoms. Such stories illustrate just how much time and energy the microarray test could save.