Before we consider the value of testing and how best to use the different types of tests available, we need to understand what they are, and their pros and cons.
Please note that it is the SARS-COV-2 virus (SARS2) that causes the COVID-19 disease.
It’s also worth mentioning my background includes work within a biotech, in an industry where initial optimism has often been followed by disappointment. So whilst I hope for the best with each of these initiatives, my professional realism is bound to surface. Until every hurdle has been jumped, nothing is certain. But we keep fingers and toes crossed for each new initiative!
TYPES OF TESTING
There are fundamentally two types of testing of individuals:
- “Got it” tests as to whether someone has the SARS2 virus, typically by looking for the RNA from within the virus
- “Had it” tests as to whether there is evidence of an immune response, typically IgG antibodies
ACCURACY OF TESTING
For any test, there are two possible error types:
- Sensitivity concerns the false negative rate and is expressed as a percentage: out of 100 truly positive patients. e.g. 30% indicates the proportion of people who should have tested positive who didn’t
- Specificity concerns the false positive rate and is expressed as a percentage: out of 100 truly negative patients
Ordinarily both types of errors should be very small. So that if you test positive, that truly means a positive result, and if you test negative that is truly negative. COVID-19 tests are generally poor on False Negatives, which is a major problem.
‘GOT IT ‘TESTS FOR THE VIRUS
Blood tests are not possible for the SARS-COV-2 virus's RNA as initially people with COVID-19 do not have the virus in their blood.
COVID-19 usually starts as a respiratory disease. So tests usually involve taking a sample of cells from the or nose and/or throat, and potentially mouth.
RT-PCR is the main test type currently used in the UK, which looks for the viral RNA.
This sounds complex, and so it is. It is certainly not as simple as a litmus paper test or a blood test. Expansion of testing capacity has therefore been expensive and slow.
For RT-PCR tests, a swab of material is taken from the back of the throat (Lower Respiratory Tract) and nose (Upper Respiratory Tract), which is then subject to testing.
Here’s a description of how it works under the sub-heading “How does real time RT–PCR work with the COVID-19 virus?” Complex equipment is needed, which heats and cools the sample repetitively, and takes a minimum of 3 hours per test.
A variety of manufacturers around the world have developed different forms of RT-PCR tests, which can be multi-step or a single step.
There are also a variety of test kits, being the swabs sent out to car homes and households to take swab samples. There have been problems woth Randox kits, resulting in at least 750,000 being recalled.
For each commercial test, a positive result is reliably correct. ONS say better than 0.04% false for the tests used in the UK. A review of academic papers on RT-PCR testing suggests this time profile after infection, given testing has been focused on people with symptoms. Lower
Respiratory Tract is for tests using throat swabs, Upper RT is for nose swabs:
RT-PCR testing will also pick up viruses before someone displays symptoms. It is reasonable to assume that would be from around the time someone becomes infectious, perhaps a couple of days after they acquire the virus. As yet we don’t know.
RT-PCR is looking for viral RNA, so it cannot distinguish active viruses from viral remnants. ‘False Positives’ can arise because remnants can be coughed up from the lungs into the lower airway, after someone is no longer infectious, Therefore there will tend to be a longer period of positive tests than the person is infectious. Remnants may also explain the difference in the two lines in the graph above.
But these issues are nothing compared to the inherent issues suffered by all the RT-PCR tests:
- False negatives risk releasing infectious people with symptoms out into the community, who would otherwise have self-isolated along with their household!
- The false negative rate is some 30% if swabs are taken professionally, which means the test often has to be repeated when someone is clearly symptomatic.
- Self-swabbing is much less reliable, which is why England has now dropped that method except under supervision [video link to follow]
- SARS-COV-2 has an ever-growing number of variants, already well over 100. It is a challenge to design tests that pick up all variants whilst avoiding other less harmful coronaviruses, such as some common colds
- Processing each test isn’t quick. Each test takes anything over three hours, though laboratories take on average between six and eight hours. Turnaround time for each test is long, mostly within 24 hours, but around a quarter of tests are up to a couple of days, We need near-instant.
- Each test is expensive. It is expensive to collect samples, it needs expensive reagents and expensive analysis equipment which has to be used by expensive staff in expensive laboratories.
- Having to regularly repeat the test to cover false negatives makes the cost of testing each person horrendous
As a result, it has taken some time to build up testing capacity in the UK, and testing in the community has been restricted only to people who are displaying COVID-19 symptoms. But that misses people who are non-symptomatic – either not yet symptomatic (pre-symptomatic) or who have the virus but never display the primary symptoms (asymptomatic).
Having said that, they are now allowing non-symptomatic people to be tested in hotspots such as Leicester. Here is the NHS page that sets out the people who are eligible. Indeed if you say you don’t have symptoms, it doesn’t seem to stop you booking a test, wherever you are.
It takes 2-14 days from catching the virus before initial symptoms, and typically 5 days. People are infectious during this pre-symptomatic period. People who have developed symptoms will usually self-isolate without needing to be tested, so the objective must be to use testing to find non-symptomatic people and get them out of circulation too. The current testing strategy in the general community, outside of hospitals and other institutions, doesn’t officially look for such people, except in those hotspots. Indeed, it's only in hotspots that mass testing of everyone is practicable and cost.-effective.
In an ideal world, every man woman and child should be tested frequently until (and if) they show a positive test. Children can be asymptomatic carriers. Ideally everyone tested every day, but if we say once every 5 days, the 55 million population of England would need 11 million tests a day. Clearly impractical, and obscenely expensive. There’s no point just asking for more tests, the testing strategy must be more focused.
That’s using RT-PCR more selectively, and preferably with better types of tests. Any alternative ‘Got it’ tests must be;
- Simpler
- Quicker
- Cheaper
- Preferably more reliable, by not being based on swabbing
- Yet clearly identify all variants of SARS-COV-2 without counting other coronaviruses
Update 9/8/20
Using a simpler 'NudgeBox (left) from an Imperial College London spinout company called DNANudge, the test offers results in less than 90 minutes. It still uses swabbing, but has a way to assess how good the swab is thereby reducing false negatives to below 6% instead of over 30%.
This is one of two tests just announced as being rolled out by the NHS across hospitals and care homes, the other being LamPORE based on different technology described below.
Reverse Transcription Loop-mediated isothermal AMPlification (RT-LAMP)
RT-LAMP sounds as complex as RT-PCR, but RT-LAMP is quicker and easier because complex equipment to change temperatures is not needed. Specific ‘primers’ need to be produced to be compatible with the viral RNA. Again variants in the RNA can potentially affect sensitivity.
The company behind the LamPORE tests, Oxford Nanopore, has provided extensive information about its role in battling COVID-19:
- Previously sequencing the RNA in the virus SARS-COV-2
- The new LamPORE tests
Whilst the press release is upbeat, the impression from the LamPORE page suggests the test is still in development, especially for alternatives to swabbing, and this initial use is part of trialling it more widely. “Status at 30 July 2020…. We will keep you informed about the rapid progress of its development here, with future regulatory updates.” So we’ll have to see.
Oxsed believes the test can produce results in 30-45 minutes, and can be used without specialist skills or equipment. It is therefore potentially suitable for use in community care, schools, airports or even for home self-testing. Though does still rely on swabbing, with its likelihood of false negatives.
Here's an interview with the team behind Oxsed's test, providing further details.
At £20 a test, it is considerably cheaper than the tests above. Which just goes to show how expensive the current testing must be. It also avoids the problem with the other tests above for which the test reagents are increasingly difficult to source. That overcomes another barrier to expanding testing capacity for the SARS-COV-2 virus.
With swabbing and self-swabbing being such an issue, you have to wonder why saliva isn’t simply used. Presumably because it isn’t a reliable way to collect enough RNA. A 4-week trial using saliva was announced as starting on 22 June in Southampton. This is using RT-LAMP and compared to swabbing using RT-PCR. As yet no news, so presumably the trial has only confirmed that saliva isn’t a suitable method. But as ever, fingers crossed.
There are presumably all manner of other initiatives around the world to harness the relative simplicity of RT-LAMP. Maybe one will take the world by storm, even cheaper and convenient than Oxsed’s. Perhaps it will use a completely different technology. Maybe another from the ‘Oxford hothouse’.
Sniffer Dogs
The idea is to deploy dogs at airports and other ports of entry into a country, a key requirement for keeping infections zero or low.
However they may miss people in the early stages of infection before symptoms. Hopefully dogs will be able to pick up a scent by the time people are infectious, if not before. Fingers crossed!
By the looks on their faces, the first dogs are high on it already!
The trial is being extended by the London School of Hygiene and Tropical Medicine. They say "The research team are aiming to recruit thousands of people in England, who have mild COVID-19 symptoms and are due to have a swab test, or have had a swab test conducted in the previous 24 hours.".
Artificial Intelligence
Oxford has also produced this Artificial Intelligence (AI) test that can produce a rapid result in under an hour from the basic tests routinely carried out when someone is admitted to hospital It will be interesting to see how this develops. Unfortunately the physical tests needed would prevent it from home use, or in any other non-hospital setting..
“HAD IT” TESTS
When the body’s immune system fights off the virus, some of the items which have been produced hang around. Testing for those that are specific to the SARS-COV-2 virus will identify who has had the COVID-19 disease.
When the body defends against the SARS-COV-2 virus, it does so by creating a number of immune responses which all involve proteins, such as to bind onto the spike proteins of the virus to de-activate it. By testing for one or more of these responses in the blood, by taking a blood sample, the patient can be tested as to whether they have had the virus.
They cannot usually act as “Got it” tests, as positive results typically only arise something like 14 days after infection, after someone is no longer infectious.
The human immune system is complex, and this is a simplification for purposes of this topic of testing. I am indebted to a Professor of Immunology here in Oxford for help in compiling this summary:
- B lymphocyte cells are white blood cells produced in the Bone marrow that generate antibodies. These are Y shaped proteins circulating in the blood that vary greatly, each specifically recognising a particular foreign target (eg, viral spike or bacterial coat), which it can neutralise, clump or help to dispose of:
- The first defence is the IgM antibody, which is a large, complex antibody with 5 Y shaped protein structures. It usually starts to appear 3-5 days after an infection. It recruits 'complementary proteins' that kill bacteria. Its levels tend to settle to lower levels after 2-3 weeks. Its specificity for its targets is not as exacting as the more refined IgA and IgG antibodies that take over. If the level of IgM is raised, it does not indicate which type of virus is present, IgM is not used for testing.
- IgA antibodies are produced by about day 7, and can be single or double Ys. They are specialised to protect mucosal surfaces (mouth, throat, airways, gut) against local intruders, such as bacteria or influenza viruses. The levels stays detectable for longer than IgM, but not long enough to be suitable for testing.
- Smaller IgG antibodies are produced, which are single Ys, appearing after 7-10 days. They level tends to increase during the next 2 weeks or so, and the antibodies evolve to improve their binding strength for their particular virus spike. IgGs in an individual are usually a complicated mix of slightly different products of different B cell clones, and some may also recognise related viral variants better than others They tend to persist for several weeks, or many months if they have been evoked by a severe infection or a replicating viral vaccine. This persistence and being specific for a particular virus, means IgG is used as the basis for antibody testing. Though in terms of avoiding false negatives, there is the complexity of having to detect IgG variants, both within an individual and the different viral variants across a population.
- Here's an indicative graph, from early in the pandemic, of IgM and IgG production compared to the level of the SARS2 virus
- Subsequent research has shown that IgM, IgA and IgG appear more quickly for COVID-19 than would normally be expected, as shown in this graph. This gives a more reliable estimate of persistence of each type of antibody, which are all significant for more than 7 weeks;
- T-cells are lymphocyte cells circulating in the blood, produced in the Thymus gland. They have not been used yet for testing, but are described here because of their general importance. There are a variety of T-cells types performing a variety of important roles:
- Antibody-producing B cells need to be switched on by their preferred target plus an additional stimulus from 'Helper T cells'. These use antibody-like proteins confined to their surfaces.
- 'Killer T-cells' with proteins on their surface can recognise and kill virus-infected cells, so creating a 'fire-break' that stops the infection from spreading.
- Both types can produce a variety of 'cytokines', which are immune protein 'hormones' that circulate in the blood. Many of these stimulate inflammation, which helps to recruit a variety of cells and proteins that help to clear away infections and the resulting debris. These are normally balanced by other 'calming' cytokines that restrain inflammation. Unfortunately, immune responses vary a lot between individuals. Any upset to this balance can lead to inflammatory 'cytokine storms' that make a minority of subjects very ill. This is a major cause, but not the only cause, of hospitalisation and death from COVID-19.
Here's a more detailed description of many facets of the SARS-COV-2 virus and how the body responds to it
Antibody tests are being conducted as part of the NHS Test and Trace service, but only by invitation. This is typically frontline staff. There is also sample testing such as being carried out by the ONS.
Again a variety of manufacturers are producing antibody tests. Each is having a problem with sensitivity (False Negatives), erroneously telling someone they haven’t had the virus, partly due to if the test is taken too early after infection. This issue is holding back large scale roll-out of antibody testing by the NHS.
Nonetheless antibody tests are now available commercially, such as from BUPA for £65. This appears to be open to anyone. BUPA says “Our tests have been evaluated for their accuracy by Public Health England. The specificity [False Positives] of our tests was evaluated as being over 99%. The sensitivity [False Negatives] of our tests was evaluated as being between 87-93%. No test is 100% accurate and a negative test therefore does not rule out previous COVID-19 infection.”
One of the issues is that BUPA says antibodies “are best detected over 14 days after infection began”, resulting in false negatives before then.
Other antibody tests are available if you do a search. This from Lloyds Chemists is a kit sent to your home for £59, which is then sent off to a laboratory. Though they are currently out of stock.
No doubt a whole range of other tests for antibodies and other immune responses are being developed across the globe, and prices will hopefully reduce. Watch this space.
IMMUNITY
From experience with other coronaviruses, we might expect that a person who has had the virus will have developed an immune response that will grant “natural immunity” for some period of months at least.
Sweden’s testing service is as yet to record the same person returning with a second COVID-19 illness, according to Anders Tegnell, their state epidemiologist [video to follow]. That sounds promising for immunity, but these are early days. It is too early to tell:
- How long immunity might last. Six months is suspected
- Whether mild illness is enough to provide immunity.
- With some other diseases, a second illness can be worse than the first. It is not known if COVID-19 has this effect
A Sting in the Immunity Tail
Until vaccination is widely available, there is talk of providing “Immunity Passports” for people who have been tested positive for COVID-19, either as a result of a “Got it” test or a “Had it” test.
Such people could be exempted from any restrictions, which is obviously good news for them. That is likely to translate into work opportunities not available to others, including the self-employed and those on zero hours contracts. An immunity passport would likely become a condition of getting or retaining a job.
Conversely that puts everyone who hasn’t yet had COVID-19 into a difficult position. Do they risk catching the virus?:
- With the likelihood they will initially be ill for 2-3 weeks
- The risk of being hospitalised for longer
- The risk of dying
- The risk of LongCOVID which could leave them out of action for much longer, potentially in great pain
You can just imagine more “COVID-19 parties” to catch the virus, with a consequent surge of cases overwhelming the NHS.
Maybe it will be impossible to stop this all. A real challenge for government policy makers and law makers, and then law enforcement.
VACCINATION
A host of vaccines are being developed around the globe.
One of the leaders is another Oxford development, kniwn as the Oxford vaccine. Commercially this is a joint venture between Oxford University’s Jenner Institute, a spin-out company called Vaccitech for some of the commercialisation, and AstraZeneca for scaling up production of the vaccine.
A production facility is already being built just outside Oxford, ready for when the vaccine is approved, hopefully early 2021. This will also be available for other vaccine products as a broader Vaccines Manufacturing Innovation Centre (VMIC).
Early trials of the Oxford vaccine have uncovered no safety concerns, and two immune system responses have been confirmed:
- A T cell response within 14 days of vaccination (white blood cells that can attack cells infected with the SARS-CoV-2 virus)
- An antibody response within 28 days
Many more hurdles to cross, though, as the trials continue:
- Safety, such as ensuring a ‘cytokine storm’ is never produced.
- Effectiveness, such as for how long? Boosters every year, or more regularly?
COMMUNITY TESTING
There is evidence that individuals as young as 10 have viral RNA remnants on this faecal stools. As yet this has not been made into a "Got it" test. But viral RNA can be detected in sewerage. The idea has been around for some time.
Studies earlier this year indicated that viral RNA appears in sewerage 2-3 weeks before hospitalisations in that area.
Professor Davey Jones from Bangor University says "All the evidence suggests that we can potentially see a signal in wastewater before we see a spike in infections in the community."
An early warning of any local surges. That's just what we need. Testing of individuals can then be focused to those areas.
NEXT STEPS
Having identified the available tests and their pros and cons, we now need to consider how best to use them.
No comments:
Post a Comment