There are 4 non-mutually exclusive reasons for this and the sample size is 1000 in Manus Brazil.
*First, the SARS-CoV-2 attack rate could have been overestimated during the first wave, and the population remained below the herd immunity threshold until the beginning of December, 2020. In this scenario, the resurgence could be explained by greater mixing of infected and susceptible individuals during December. The 76% estimate of past infection might have been biased upwards due to adjustments to the observed 52·5% (95% CI 47·6–57·5) seroprevalence in June, 2020, to account for antibody waning…and the mandatory exclusion of donors with symptoms of COVID-19 is expected to underestimate the true population exposure to the virus. *
Second, immunity against infection might have already begun to wane by December, 2020, because of a general decrease in immune protection against SARS-CoV-2 after a first exposure. Waning of anti-nucleocapsid IgG antibody titres observed in blood donors might reflect a loss of immune protection, although immunity to SARS-CoV-2 depends on a combination of B-cell and T-cell responses.
A study of UK health-care workers showed that reinfection with SARS-CoV-2 is uncommon up to 6 months after the primary infection. However, most of the SARS-CoV-2 infections in Manaus occurred 7–8 months before the resurgence in January, 2021; this is longer than the period covered by the UK study, but nonetheless suggests that waning immunity alone is unlikely to fully explain the recent resurgence. Moreover, population mobility in Manaus decreased from mid-November, 2020, with a sharp reduction in late December, 2020,
suggesting that behavioural change does not account for the resurgence of hospitalisations.
Third, SARS-CoV-2 lineages might evade immunity generated in response to previous infection.
Three recently detected SARS-CoV-2 lineages (B.1.1.7, B.1.351, and P.1), are unusually divergent and each possesses a unique constellation of mutations of potential biological importance.
Of these, two are circulating in Brazil (B.1.1.7 and P.1) and one (P.1) was detected in Manaus on Jan 12, 2021.
P.2 variants with the E484K mutation have been detected in two people who have been reinfected with SARS-CoV-2 in Brazil,
and there is in-vitro evidence that the presence of the E484K mutation reduces neutralisation by polyclonal antibodies in convalescent sera.
Fourth, SARS-CoV-2 lineages circulating in the second wave might have higher inherent transmissibility than pre-existing lineages circulating in Manaus. The P.1 lineage was first discovered in Manaus.
In a preliminary study, this lineage reached a high frequency (42%, 13 of 31) among genome samples obtained from COVID-19 cases in December, 2020, but was absent in 26 samples collected in Manaus between March and November, 2020.
Thus far, little is known about the transmissibility of the P.1 lineage, but it shares several independently acquired mutations with the B.1.1.7 (N501Y) and the B.1.325 (K417N/T, E484K, N501Y) lineages circulating in the UK and South Africa, which seem to have increased transmissibility.
Contact tracing and outbreak investigation data are needed to better understand relative transmissibility of this lineage.