Repeated SARS-CoV-2 Introductions with Limited Local Establishment in Bangladesh, 2020-2025

Viruses leave a trail of genetic changes as they move through a population, much like footprints in the snow. When scientists sequence the genetic code of a virus, they can read these changes to reconstruct where the virus came from, how it traveled, and whether it started spreading locally or arrived from elsewhere. This field of study, known as genomic surveillance, has been incredibly successful in wealthy nations where testing and sequencing happen constantly. However, a major question remained unanswered: can we trace these viral journeys in countries where sequencing is rare and happens only occasionally? In places like Bangladesh, where resources are limited and data is sparse, it was unclear if scientists could reliably distinguish between a virus that simply passed through and one that took root and spread. Understanding this difference is vital for public health, because stopping a virus requires knowing whether to focus on border controls to block new arrivals or on local measures to stop community spread.

To answer this, researchers gathered virus genomes collected in Bangladesh between March 2020 and July 2025, comparing them against a global background of viral sequences. They built a family tree of the virus to map out when and where different lineages, or branches of the virus, appeared in the country. By using statistical models to account for the fact that they only saw a small fraction of the total infections, they could infer how many times the virus was introduced from outside. The study found that the virus entered Bangladesh 361 separate times. More than half of these introductions, specifically 51.8 percent, resulted in only a single sampled genome with no evidence of the virus spreading further to others. This suggests that many viral arrivals died out quickly without establishing a local chain of transmission.

The researchers then looked at where these viruses came from. They found that India and the Gulf region were the dominant sources, bringing in significantly more virus lineages than would be expected by chance. Other regions, including East Asia and South-East Asia, contributed far less than the statistical models predicted. The team also examined specific genetic changes, or mutations, that appeared in the Bangladeshi viruses. They distinguished between mutations that the virus brought with it from abroad and those that arose locally after the virus entered the country. They identified eighteen mutations that were enriched in the local samples, which grouped into eleven independent signals. Of the 171 times these mutations appeared in a lineage, 53 of them happened after the lineage had already entered Bangladesh, proving that local evolution was occurring.

One specific mutation, known as G446V in the spike protein, provided a striking example of local evolution. This change appeared locally in 17 out of 19 different viral lineages that carried it. In one imported Delta lineage, this locally arising mutation spread to 267 of the 291 genomes, covering 91.8 percent of that group. This showed that a mutation arising within the country could become dominant even if the virus lineage itself was imported. However, the researchers noted a limitation in their ability to track trends over time. Because the amount of sequencing effort changed throughout the years, they could not determine if the country became less dependent on new virus imports as time went on. The data showed that importation decided which viral lineages circulated in Bangladesh, while local evolution determined what happened within those lineages. Although local mutations spread easily, they rarely reached the frequency of an entire lineage because the lineages carrying them were observed for only a short time before disappearing from the genetic record.