BY October last year, at least three quarters of people living in the Brazilian city of Manaus were testing positive for Covid antibodies.

The novel coronavirus had swept through the region's population leading to one of the highest infection rates in the world, partly as a result of Brazilian President Jair Bolsonaro's denial that the pandemic posed a major threat and total failure to finance and coordinate a public health response.

The blood samples, reported in the Lancet, were explicit proof that 76% of residents had been exposed to the virus - but it also led to an assumption that they would be largely immune to a second wave.

Manaus was an interesting test case for those who advocated a herd immunity approach.

HeraldScotland: From 'Resurgence of COVID-19 in Manaus, Brazil, despite high seroprevalence', Lancet, February 6From 'Resurgence of COVID-19 in Manaus, Brazil, despite high seroprevalence', Lancet, February 6

Hospitalisations for Covid had remained stable and fairly low for seven months from May to November, following the peak of the epidemic in April when authorities were forced to carve out mass graves to bury the dead.

It came as a shock therefore - 210,000 deaths later - when the number of patients being admitted to hospital with the disease suddenly started soaring again in January and, at first, no one knew why.

Had antibody prevalence been overestimated? Was immunity waning already?

We know now that what was really driving the upsurge was a new mutant strain - P.1 - which was first detected in Manaus on January 12 but had begun circulating there in December.

By the end of January, 200 bodies a day were being interred in Manaus compared to the usual 40, some hospitals ran out of oxygen, and intensive care units were overwhelmed.

READ MORE: Mystery person with Brazil variant found 

Then, on January 29, three oil workers boarded a plane from Brazil to Aberdeen, via Paris and London.

They were employees of Bureau Veritas, a global testing a certification company who provide services to the offshore industry, and had been working in Brazil since December 23.

All three had tested negative for coronavirus in the 72 hours prior to departure, a mandatory requirement for all international arrivals into the UK.

HeraldScotland: Brazil has the second highest coronavirus death toll after the US, at more than a quarter of a millionBrazil has the second highest coronavirus death toll after the US, at more than a quarter of a million

If they had landed at Heathrow any time from February 15 onwards, all three would have been whisked away into supervised hotel quarantine for 10 days.

But in late January the 'red list' country rule was not yet in operation and instead they flew on to Scotland, where they self-isolated on arrival.

On February 5, all three tested positive for coronavirus, with two subsequently admitted to hospital - though both have since recovered.

Genomic testing would later reveal that their infections were caused by the P.1 variant, which had not previously been detected in the UK.

This suggested they must have been incubating the virus at the time they boarded the plane from Brazil.

The good news is that, to date, there is no evidence that the P.1 variant has escaped into the community.

HeraldScotland: Three oil workers tested positive on February 5, while self-isolating in AberdeenThree oil workers tested positive on February 5, while self-isolating in Aberdeen

More worrying, however, is that one in three of the passengers who shared their London to Aberdeen flight were yet to be traced as of Wednesday (this has been blamed on invalid contact details which suggests, somewhat bizarrely, that a very high number of air travellers are providing duff details to contact tracers).

Meanwhile, in England - where three P.1 infections were also confirmed last weekend - the mystery third case has finally been tracked down in Croydon, near London.

Two of the three had already been confirmed as members of the same South Gloucestershire household with recent travel to Brazil, but the third was unlinked - and potentially not self-isolating.

READ MORE: NHS bosses told to offer vaccinators shifts 'on same basis' amid pay row

All that was known was that they had taken a Covid test on February 12 or 13 and had not been given their results because they did not complete their test registration card correctly.

Yesterday, Health Secretary Matt Hancock confirmed that Patient X has now been identified and had "stayed at home".

There was no sign of community transmission, he said, but local testing is being rolled out.

HeraldScotland: A total of 3,566 positive cases have been reported in the past seven days, the lowest since late SeptemberA total of 3,566 positive cases have been reported in the past seven days, the lowest since late September

The confirmation of the P.1 variant in the UK this week has coincided with an alarming study into the Manaus outbreak by Imperial College London - although the researchers caution that its findings "should not be generalised to other epidemiological contexts" where the strain has been found.

So far, it has been detected in 25 countries including Japan, Italy, South Korea, the United States, India, Germany, France and the Netherlands.

The ICL study found that the proportion of Covid infections caused by the P.1 variant went from zero to 87 per cent in the space of eight weeks.

They estimate that it is 1.4 to 2.2 times more transmissible than the original Wuhan strain thanks to a "unique constellation of mutations".

For comparison, the UK 'Kent' variant is thought to be around 50% - 0.5 times - more infectious.

The Brazil variant's mutations include N501Y (which the Kent strain also shares, allowing it to spread more easily), but also one called E484K (shared by the South African variant) which has reshaped the spike protein and means that antibodies are unable to bind to them as effectively.

READ MORE: Geneticists seek blood samples from people who had mild Covid symptoms

This raises the prospect that it could evade vaccines and also explains why herd immunity did not work in Manaus: people were being reinfected.

Dr Nuno Faria, reader in viral evolution at ICL and lead author of the study, told a press briefing on Tuesday: “If 100 people were infected in Manaus last year, somewhere between 25 and 61 of them are susceptible to reinfection with P.1.”

The study has not yet been peer-reviewed and Professor Sharon Peacock, head of the UK’s genetic surveillance programme, stressed that it was too soon to "speculate on vaccine efficacy”.

Pfizer and Moderna are already developing booster shots against the South African variant which would hopefully show similar effectiveness against the Brazil variant.

Nonetheless, the experience in Manaus should serve as a warning against complacency.

In the UK, where the Kent variant accounts for more than 85% of cases, vaccines are driving down deaths and serious illness and potentially even slowing transmission.

The risk of a P.1-type variant - whether imported or able to evolve domestically if we lift restrictions too quickly and allow infections to surge - is a bit like an athlete slowing down too close to the finish line, only to be overtaken.

And that would be a bitter blow indeed.