Early transmission of SARS-CoV-2 in South Africa: An epidemiological and phylogenetic report

Abstract

Background: The emergence of a novel coronavirus, SARS-CoV-2, in December 2019, progressed to become a world pandemic in a few months and reached South Africa at the beginning of March. To investigate introduction and understand the early transmission dynamics of the virus, we formed the South African Network for Genomics Surveillance of COVID (SANGS_COVID), a network of ten government and university laboratories. Here, we present the first results of this effort, which is a molecular epidemiological study of the first twenty-one SARS-CoV-2 whole genomes sampled in the first port of entry, KwaZulu-Natal (KZN), during the first month of the epidemic. By combining this with calculations of the effective reproduction number (R), we aim to shed light on the patterns of infections that define the epidemic in South Africa.

Methods: R was calculated using positive cases and deaths from reports provided by the four major provinces. Molecular epidemiology investigation involved sequencing viral genomes from patients in KZN using ARCTIC protocols and assembling whole genomes using meticulous alignment methods. Phylogenetic analysis was performed using maximum likelihood (ML) and Bayesian trees, lineage classification and molecular clock calculations.

Findings: The epidemic in South Africa has been very heterogeneous. Two of the largest provinces, Gauteng, home of the two large metropolis Johannesburg and Pretoria, and KwaZulu-Natal, home of the third largest city in the country Durban, had a slow growth rate on the number of detected cases. Whereas, Western Cape, home of Cape Town, and the Eastern Cape provinces the epidemic is spreading fast. Our estimates of transmission potential for South Africa suggest a decreasing transmission potential towards R=1 since the first cases and deaths have been reported. However, between 06 May and 18 May 2020, we estimate that R was on average 1.39 (1.04 - 2.15, 95% CI). We also demonstrate that early transmission in KZN, and most probably in all main regions of SA, was associated with multiple international introductions and dominated by lineages B1 and B. The study also provides evidence for locally acquired infections in a hospital in Durban within the first month of the epidemic, which inflated early mortality in KZN.

Interpretation: This first report of SANGS_COVID consortium focuses on understanding the epidemic heterogeneity and introduction of SARS-CoV-2 strains in the first month of the epidemic in South Africa. The early introduction of SARS-CoV-2 in KZN included caused a localized outbreak in a hospital, provides potential explanations for the initially high death rates in the province. The current high rate of transmission of COVID-19 in the Western Cape and Eastern Cape highlights the crucial need to strength local genomic surveillance in South Africa.

Funding: UKZN Flagship Program entitled: Afrocentric Precision Approach to Control Health Epidemic, by a research Flagship grant from the South African Medical Research Council (MRC-RFA-UFSP-01-2013/UKZN HIVEPI, by the the Technology Innovation Agency and the the Department of Science and Innovation and by National Human Genome Re- search Institute of the National Institutes of Health under Award Number U24HG006941. H3ABioNet is an initiative of the Human Health and Heredity in Africa Consortium (H3Africa).

Figure 1:

Timeline of measures implemented in South Africa from the first detected COVID-19 case on 5 March to the expansion of the lockdown in April 2020.

Figure 2:. Summary of the COVID-19 epidemic in South Africa.

A) Numbers of COVID-19 cases and deaths in the Western Cape (WC), Gauteng (GP), KwaZulu-Natal (KZN) and the Eastern Cape (EC). B) Geographic map showing the location of South African provinces.

Figure 3:. Phylogenetic analysis.

(A) A time scaled Maximum likelihood tree of 1849 sequences including 21 genotypes from KwaZulu-Natal, South Africa. Major lineages of SARS-CoV-2 are labelled. B) Monophyletic cluster of KZN sequences. (C) Stacked barplot showing the lineage breakdown of the dataset by region.

Figure 4:. Time stamped phylogenetic trees of the three lineages of SARS-CoV-2 found in KwaZulu-Natal (KZN).

The genomes produced in this study are marked with a red triangle, and the NICD genome by a red square. The geographic region of the other sequences is marked with coloured circles.