Hancock needs to make antibody testing his next big project

After much excitement in mid-to-late March among the British authorities about antibody tests for Covid-19, the UK has still not managed to roll these out in significant numbers. The reasons for this are related to the complexities of developing a point-of-care testing device for antibodies as much as a failure to mobilise private sector companies with greater urgency. Yet there are also some hopeful signs for the future.

The British government has now managed to launch a large-scale community testing programme, which will include lab-based antibody tests, using a process known as ELISA, to test 1,000 households across the country. Yet what is still missing is the remote form test – “the lateral flow immunoassay” (LFIA) – which will allow people to use a “dipstick” or “pregnancy-style” testing kit for antibodies in GP surgeries or in the comfort of their own homes.

Developing an effective remote, point-of-care test kit, using the LFIA, is important, experts agree, because it is far quicker than the lab-based ELISA test. The technology is also relatively cheap and easy to manufacture, potentially allowing for a rapid scaling up of antibody tests and their widespread circulation throughout the community.

This was the plan, and the government was initially hoping to get large-scale antibody testing underway much sooner. But these efforts have now stalled, even if there are encouraging signs that progress is being made, both by the public authorities and by private companies, behind the scenes.

On 18 March, Sir Patrick Vallance said that Public Health England’s antibody test was “progressing very fast”, with a prototype “in house” and that the health authorities were “looking at ways to get a much more widespread version of that out.”

Then, on 25 March, Sharon Peacock, the Director of the National Infection Service at PHE, announced to the House of Commons Science and Technology Committee that millions of these could potentially be ready “within days”.  However, Peacock did also say this would be conditional on PHE being “assured that they do work.”

The Prime Minister and the Health Secretary, Matt Hancock, were also enthusiastic about the prospect of a new pregnancy-style antibody test in late March, building anticipation of the imminent arrival of the new devices.

It was even suggested that these could be ordered and delivered by big-name distributors and pharmacies, such as Amazon and Boots, generating an effective exit from the UK’s national lockdown.

The first blow to these plans came in early April, when the government admitted that none of the millions of antibody tests it had ordered from China had met the accuracy standards required by PHE and the government’s testing guru, Professor John Newton.

Then another roadblock arose, after it emerged on April 21 that nine different commercially-available point-of-care LFIAs had also failed to meet PHE’s exacting standards for accuracy. In the formal report on these antibody tests, published by the National Testing Scientific Advisory Panel, it was concluded that the LFIAs were not yet precise enough for public use.

The study of these commercially-available antibody tests had specific criteria: they had to have a “specificity” and “sensitivity” of no less than 98%.

Sensitivity means the proportion of true positives picked up by the test. That is, when people who have antibodies are correctly identified as having them. Specificity, on the other hand, means the proportion of true negatives – when people do not have the virus, and this is successfully indicated by the test device.

When these features were tested against the “gold standard” ELISA test, it was found that while the specificities of some of the LFIA tests met the standard, with a range of 100% to 95%, its sensitivities were rather less impressive, ranging from around 70% down to 55%.

The developers of antibody tests are now back at the drawing board, seeking to enhance accuracy to match the requirements of PHE.

Amid all of this, there has also been a tension between private companies, which are seeking to rapidly validate their antibody tests, and the public health authorities, which have been understandably cautious about prematurely deploying tests of which the quality cannot be assured.

One lobbyist for a diagnostics company, which has produced its own remote Covid-19 antibody test, told me that efforts to engage the private sector had been hampered by institutional inertia. He added that the difficulties of developing an accurate antibody test have been compounded by an over-centralisation of the process of validation and a lack of support for commercial companies seeking to provide their expertise.

“Their cultural, innate reaction is to do everything for themselves, and not to be sullied by the private sector”, the lobbyist said. He sees this as a “cultural issue within PHE in particular”, expressing concern that “they always want the best rather than the good, demanding a fine degree of certainty at the expense of taking the urgent action and initiative required.”

It is pointed out by others that even the supposed “gold standard” ELISA test, according to the National Testing Scientific Advisory Panel report, could only guarantee 85% sensitivity in reporting whether or not people have immunity.

Another diagnostics company working on LFIAs, BioSure, has also voiced its readiness to roll out a more accurate antibody test. The company has developed a 15-minute self-test which they claim to have 99.5% accuracy

But the matter is complex, with scientific and medical experts themselves divided over the right balance to be struck between institutional caution and speed. Dr Simon Clarke, Associate Professor of Cellular Microbiology at the University of Reading, warns that it is necessary to “underscore the importance of independent verification of new technologies before the claims of their manufacturers are taken at face value.”

Reaction has been told that BioSure’s own validation tests – which produced the 99.5% figure – were based upon private trials of 60 patients across 2 UK hospitals, a small initial sample size which now needs to be expanded.

In achieving this goal, more support from the government will be needed. Many smaller diagnostics companies have voiced their frustration that there have not been greater efforts made by the government and the Department of Social Care in helping them obtain the accuracy required of antibody tests. Many of them have only fairly recently been provided with the information needed to get their LFIAs up to official standards. Some firms say they were only notified of the specifications for self-test antibody tests by the Medicines and Healthcare products Regulator Agency (MHRA) and the DHSC on 24 April.

Companies also need more support in getting the biobanks of blood samples, taken from those who have recovered from Covid-19, in order to validate their kits.

An email sent out to companies from the DHSC in late April, seen by Reaction, explains that the government is “working through” the offers for help from private companies. It adds that examining private companies’ test kits “is taking time as many are untested for COVID-19 and/or are components that need to be tested for compatibility with other lab processes and components that we have already set up.”

In defence of the health authorities and the government, there are those who see the cautious role of PHE stemming not from stubborn inertia but from a caution. They are, after all, responsible for the health of British citizens. They must necessarily take a steadier approach in assuring quality control, placing their public responsibility above understandable urges for swifter action.

Stephen Evans, Professor of Pharmacoepidemiology at the London School of Hygiene and Tropical Medicine, for instance, believes that the government has “sensibly” opted for a minimum value of 98%. He explains that “If the test is to be used to show that a (hopefully) large percentage of the population is immune, enabling people to return to work, especially in healthcare, it is important that it does not give a positive result when someone does not have the antibodies and would therefore definitely not be immune.”

This would have potentially devastating consequences, he adds, because such individuals “could go into situations thinking they were immune but would actually be vulnerable to infection. This would be not only bad for the individuals, it could lead to a major resurgence of the disease in the population.”

But what happens if antibody tests cannot be made to reach the exacting standards required by the health authorities’ specifications? The question is pressing, given that no immunoassay can ever guarantee 100% accuracy, all of the time.

Several scientific and medical experts continue warn that, while developing an accurate remote antibody test is important, the relationship between the presence of antibodies and longer-term immunity to Covid-19 is still being studied. The strength of immunity, and for how long it lasts after recovery from infection, are still not fully understood.

One grey area concerns whether those who are infected while being “asymptomatic” or experiencing only mild symptoms will still produce an immune response which can be picked up by antibody tests. There have been some studies of immune responses to SARS-CoV-2 in China, for instance, which have suggested that there may be “a lack of persistent antibodies following mild disease”.

Reaction asked Alexander Edwards, an immunologist and Associate Professor in Biomedical Technology at the University of Reading, whether asymptomatic cases could be picked up later by the point-of-care LFIA antibody tests.

He said: “Each individual will generate a different immune response and this is highly variable”, but added that “we would expect to see antibody responses in anyone with a significant infection”.

Ultimately, Edwards explains, “asymptomatic cases are difficult to assess, because we don’t yet know much about them. They are, for obvious reasons, not as easy to identify as those who have symptoms. There are now some decent data sets being published, and we’ll obtain a clearer, more complete picture soon.”

There are still useful ways in which even a less accurate test can be employed to combat the Covid-19 epidemic. While they may not be suitable for individual diagnosis, they could still be used to improve population surveillance and epidemiological modelling to assess the spread of the epidemic.

Sylvia Richardson, the co-chair of the Royal Statistical Society Task Force on Covid-19, for instance, has written to the MHRA to make the case for using antibody tests which meet lower standards, to be “used in the context of a powerful suite of population surveillance studies.”

In the end, there are many creative tensions at play which explain the struggles experience by the UK government to get remote, point-of-care antibody testing up and running. The current inaccuracies of the devices involved, the complications of human immunity, and a dose of institutional inertia have created challenges.

Yet vital work continues to take place in the public and private sectors. As with every element of the British government’s response to this time of crisis, the public will have to continue in hoping that the authorities continue to learn on the go and make progress. The experience so far suggests that the tension which has existed between the public and private sectors must be better harnessed to find a way of using antibody tests for a variety of policies.

Alongside other testing and tracing programmes which are now getting underway, another big push to improve and expand antibody testing could be precisely what is needed to finally begin lifting Britain’s national lockdown. The Health Secretary, Matt Hancock, buoyed by recent targets met on PCR testing, might now look to grasp this final nettle.