Spatiotemporal evolution of SARS-CoV-2 in the Bangkok metropolitan region, Thailand, 2020–2022: implications for future outbreak preparedness

Thailand experienced five waves of coronavirus disease 2019 (COVID-19) between 2020 and 2022, with the Bangkok Metropolitan Region (BMR) being at the centre of all outbreaks. The molecular evolution of the causative agent of the disease, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has previously been characterized in Thailand, but a detailed spatiotemporal analysis is still lacking. In this study, we comprehensively reviewed the development and timelines of the five COVID-19 outbreaks in Thailand and the public health responses, and also conducted a phylogenetic analysis of 27 913 SARS-CoV-2 genomes from Thailand, together with 7330 global references, to investigate the virus’s spatiotemporal evolution during 2020 and 2022, with a particular focus on the BMR. Limited cross-border transmission was observed during the first four waves in 2020 and 2021, but was common in 2022, aligning well with the timeline of change in the international travel restrictions. Within the country, viruses were mostly restricted to the BMR during the first two waves in 2020, but subsequent waves in 2021 and 2022 saw extensive nationwide transmission of the virus, consistent with the timeline of relaxation of disease control measures employed within the country. Our results also suggest frequent epidemiological connections between Thailand and neighbouring countries during 2020 and 2021 despite relatively stringent international travel controls. The overall sequencing rate of the viruses circulating in the BMR was ~0.525 %, meeting the recommended benchmark, and our analysis supports that this is sufficient for tracking of the trend of the virus burden and genetic diversity. Our findings reveal insights into the local transmission dynamics of SARS-CoV-2 in Thailand, and provide a valuable reference for planning responses to future outbreaks.

, main text).A.6 was found to be monophyletic, comprising primarily viruses from the BMR, along only a few sequences from other Asian countries, including Cambodia (2 sequences), Hongkong (2), South Korea (1), and Singapore (1) (Supplementary Figure 2).This pattern indicates that all of the A.6 BMR sequences in the dataset were sampled from a single outbreak sharing a single origin, and the outbreaks was mainly restricted to the BMR.The estimated time to most recent common ancestor (tMRCA) of the main A.6 cluster was ~2020.17(03/03/2020), about a week before the earliest sequence within the cluster (09/03/2020).However, basal to the main cluster were two A.6 sequences from Nonthaburi, which is also part of the BMR, reported in late January (27/01/2020), and late February (24/02/2020).This suggests that the virus might have already been present in the BMR by then but without being noticed.
B.1 and B.1.1 sequences, in contrast, were found to be non-monophyletic, scattering across the base of the tree.In particular, they were found to be basal to, and / or cluster with, the B.1.1.7 (Alpha) variant, and B.1.1.529and BA (Omicron) variants (Figure 3, main text).This is not surprising however, given that B.1.1.7,B.1.1.529and BA evolved from B.1 and B.1.1 variants by definition.In addition, BMR B.1 and B.1.1 sequences were found to cluster together with sequences reported from outside Thailand but without forming large clusters (Supplementary Figure 2).These sequences likely represented sporadic imported cases that Thai authorities effectively identified, quarantined, and thereby prevented them from causing significant outbreaks within the region.
While the majority of the sequences in the clusters were from the BMR, they also contained a few sequences from other areas of Thailand, as well as some from Cambodia forming a tight clade (Supplementary Figure 3).This phylogenetic pattern suggests that there were a few sporadic transmissions of the virus from the BMR to other areas of Thailand during the 2 nd wave, as well as at least one instance of virus exportation from Thailand to Cambodia.Immediately basal to the Thai B.1.36.16 cluster are B.1.36.16 sequences from Myanmar.This observation supports potential epidemiological connections between Thailand and Myanmar in the early stages of the outbreak, previously speculated based on epidemiological profiles of the virus and patients during that time [3].

B.1.1.7 -the dominant virus variant during the 3 rd wave
The B.1.1.7 (Alpha) variant was the dominant variant during the first half of the 3 rd COVID-19 wave in the BMR (Figure 2, main text), and their sequences formed a single large cluster on the tree (Supplementary Figure 4).Unlike the A.6 and B.1.36.16 variants, while the cluster still contained only a few sequences reported from outside Thailand, it contained a relatively large number of non-BMR sequences (n = 805), and they phylogenetically intermixed with the BMR sequences (n = 1,252).These results are consistent with that these B.1.1.7 sequences were collected from a single large outbreak in Thailand, and the wide-spread of the virus in the country was likely due to extensive local transmissions of the virus, rather than multiple introductions of the virus into the country.The classic Delta B.1.617.2 variant, and its two subvariants, including AY.30, and AY.85 variants, began appearing in the BMR since the middle of the 3 rd wave (Figure 2, main text).The three virus variants gradually and steadily replaced the B.1.1.7 variant, and became the predominant variants in the region, eventually causing the 4 th wave of COVID-19 in Thailand.Phylogenetic analysis revealed that AY.30, and AY.85 formed their own distinct clusters, and B.1.617.2 sequences scattered around the base of the two (Supplementary Figure 5).This pattern was expected, as AY.30 and AY.85 are by definition within the diversity of the B.1.617.2 variant.
The two clusters were not sister groups on the tree, and basal to them were different sets of non-Thai sequences, indicating independent introductions of the viruses to the country.The MRCAs of the AY.30 and AY.85 variants were dated to ~2021.058(22/01/2021), and ~2021.112(11/02/2021), respectively.This suggests that the two variants likely emerged months before causing larger outbreaks in Thailand.In addition, while the base of the cluster of AY.85 was not well resolved, we found that positioning immediately at the base of the AY.30 cluster were a few sequences from Cambodia (6), Myanmar (1), Singapore (1), and Bangladesh (1) (Supplementary Figure 5, inset), suggestive of epidemiological linkages between Thailand and these countries in the early phases of the outbreak.

BA variants -the dominant virus variants during the 5 th wave
Numerous sub Omicron variants were detected in the sequence data collected during the 5 th wave, but which could be tentatively grouped into four major variants, including BA.1, BA.2, BA.4, and BA.5 (Figure 2, main text), each forming their own major sequence phylogenetic cluster (Supplementary Figure 6).
Analysis revealed that, unlike other variants described above, both Thai and non-Thai Omicron sequences intermixed together on the tree in all four major Omicron clusters, as well as sequences from the BMR and other regions of the country (Supplementary Figure 6).This pattern supports that the 5 th wave of COVID-19 in Thailand was a result of multiple independent virus introductions, coupled with extensive local transmissions.This result, again, underscores Thailand, particularly the BMR, as a hotspot for emerging infectious diseases.