Implementation of genomic surveillance of SARS-CoV-2 in the Caribbean: Lessons learned for sustainability in resource-limited settings

Abstract

The COVID-19 pandemic highlighted the importance of global genomic surveillance to monitor the emergence and spread of SARS-CoV-2 variants and inform public health decision-making. Until December 2020 there was minimal capacity for viral genomic surveillance in most Caribbean countries. To overcome this constraint, the COVID-19: Infectious disease Molecular epidemiology for PAthogen Control & Tracking (COVID-19 IMPACT) project was implemented to establish rapid SARS-CoV-2 whole genome nanopore sequencing at The University of the West Indies (UWI) in Trinidad and Tobago (T&T) and provide needed SARS-CoV-2 sequencing services for T&T and other Caribbean Public Health Agency Member States (CMS). Using the Oxford Nanopore Technologies MinION sequencing platform and ARTIC network sequencing protocols and bioinformatics pipeline, a total of 3610 SARS-CoV-2 positive RNA samples, received from 17 CMS, were sequenced in-situ during the period December 5th 2020 to December 31st 2021. Ninety-one Pango lineages, including those of five variants of concern (VOC), were identified. Genetic analysis revealed at least 260 introductions to the CMS from other global regions. For each of the 17 CMS, the percentage of reported COVID-19 cases sequenced by the COVID-19 IMPACT laboratory ranged from 0·02% to 3·80% (median = 1·12%). Sequences submitted to GISAID by our study represented 73·3% of all SARS-CoV-2 sequences from the 17 CMS available on the database up to December 31st 2021. Increased staffing, process and infrastructural improvement over the course of the project helped reduce turnaround times for reporting to originating institutions and sequence uploads to GISAID. Insights from our genomic surveillance network in the Caribbean region directly influenced non-pharmaceutical countermeasures in the CMS countries. However, limited availability of associated surveillance and clinical data made it challenging to contextualise the observed SARS-CoV-2 diversity and evolution, highlighting the need for development of infrastructure for collecting and integrating genomic sequencing data and sample-associated metadata.

Fig 1. Movement of SARS-CoV-2 positive samples, data and sequencing results (Dec 5^th 2020 to Dec 31^st 2021).

From December 5^th 2020 to December 31^st 2021, a total of 3667 viral RNA extracts from SARS-CoV-2 positive samples were submitted to the UWI COVID-19 IMPACT laboratory for sequencing. Of those 1670 were received, via TPHL, from T&T’s five regional health authorities (NWRHA, NCRHA, ERHA, SWRHA, and TRHA) as well as from private laboratories within T&T. The remaining 1997 samples were received from CARPHA, from seventeen CARPHA Member States. Sequencing results for samples received from CARPHA were reported back to CARPHA. For the 2597 samples from T&T (regardless of whether received via TPHL or CARPHA), results were reported to the Chief Medical Officer at the T&T Ministry of Health in a weekly cumulative report, except for first time reports of VOCs and VOIs which were reported to the Chief Medical Officer immediately upon detection until local transmission was confirmed. Results for samples received from other CARPHA Member States were reported to CARPHA as soon as they were obtained and relayed to the respective member state by CARPHA. Samples were submitted for sequencing with accompanying sample data. Broken red lines indicate routes where the COVID-19 IMPACT laboratory had no control over the communication of metadata. TPHL = Trinidad Public Health Laboratory. NWRHA = North-west Regional Health Authority. NCRHA = North-central Regional Health Authority. ERHA = East Regional Health Authority. SWRHA = South-west Regional Health Authority. TRHA = Tobago Regional Health Authority. TPHL = Trinidad Public Health Laboratory. Caribbean Public Health Agency = CARPHA. VOC = Variant of Concern. VOI = Variant of Interest.

Fig 2. Variants of concern and variants of interest identified per country by the Covid-19 IMPACT sequencing laboratory as at December 31^st 2021.

From December 5^th 2020 to December 31^st 2021 a total of 3610 samples were sequenced with sufficient coverage obtained for 2975 (82·4%) to enable assignment to a Pango lineage. For each of the 17 CARPHA CMS participating in the COVID-19 IMPACT project (i.e., Anguilla, Antigua and Barbuda, Bahamas, Barbados, Bermuda, British Virgin Islands, Cayman Islands, Dominica, Grenada, Guyana, Jamaica, Montserrat, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Turks and Caicos Islands, and T&T) the percentages of VOCs (Alpha, Beta, Gamma, Delta, and Omicron) and VOIs (Epsilon, Iota, Lambda, and Mu) identified are indicated in a pie chart. All other lineages identified are grouped together and indicated as “Other”. The numbers below the pie charts indicate respectively the number of sequences and lineages identified in each country, as at December 31, 2021. CARPHA = Caribbean Public Health Agency. CMS = CARPHA Member States. VOC = Variants of Concern. VOI = Variants of Interest. Map base layer modified from Serafy et al. 2015 [28] (License CC by 4.0; available at https://doi.org/10.1371/journal.pone.0142022.g001).

Fig 3. Bayesian timescaled phylogenetic reconstruction.

SARS-CoV-2 sequences with over ≥75% genome coverage generated by the COVID-19 IMPACT laboratory between December 5th 2020 to December 31st 2021 are indicated with circles filled in red (n = 2716). Those sequences comprising the background dataset are grouped as either sequences from the Caribbean (specifically CARPHA Member States, indicated by red branches and open circles) or sequences from international countries (i.e. non-CARPHA Member States, indicated by grey branches). VOCs identified are highlighted. CARPHA = Caribbean Public Health Agency. VOC = Variant of Concern.

Fig 4. Frequency of variants of concern and interest identified by the COVID-19 IMPACT sequencing laboratory per epidemiological week for those CMS with at least 75 samples sequenced.

Each histogram shows the number of VOCs (indicated in blue) and VOIs (indicated in green) per week in each CMS. All other lineages are grouped together as “Other” (yellow). Unsuccessfully sequenced samples are not represented as sufficient genome coverage was not obtained to be able to assign a Pango lineage. The purple line represents the number of confirmed reported COVID-19 cases for the country per week based on data available from The Johns Hopkins University Center for Systems Science and Engineering COVID-19 Data Repository. CMS = CARPHA member states. VOC = Variants of Concern. VOI = Variants of Interest.

Fig 5. SARS-CoV-2 genomic surveillance by the COVID-19 IMPACT project.

(A) Percentage of cases sequenced by the COVID-19 IMPACT laboratory per epidemiological week per CMS, between January 28^th 2020 and December 31^st 2021. For each CMS, the percentage of sequenced cases per epidemiological week is indicated by a coloured rectangle. Weeks during which there were no reported cases for that CMS are indicated in dark grey, and weeks in which there were cases reported for the CMS, but none were sequenced by the COVID-19 IMPACT laboratory are indicated in light grey. The overall percentage of cases sequenced by the COVID-19 IMPACT laboratory across the entire period is indicated for each CMS in column i. The percentage of samples received by the COVID-19 IMPACT laboratory that were successfully sequenced across the entire period is indicated for each CMS in column ii. The number of reported cases for each CMS is based on data available from The Johns Hopkins University Center for Systems Science and Engineering COVID-19 Data Repository. (B) Proportion of reported cases sequenced per CMS, between January 28^th 2020 and December 31^st 2021. Each circle indicates an epidemiological week during which at least one reported case for the respective CMS was sequenced and represents the percentage of sequenced cases for that week. Weeks where there were cases reported for the CMS but none were sequenced are not shown. The diameter of each circle represents the incidence (cases per 100/habitants) in that week for the respective CMS. The overall percentage of cases sequenced by the COVID-19 IMPACT laboratory for each CMS across the entire period is indicated by a black line. The number of reported cases for each CMS is based on data available from The Johns Hopkins University Center for Systems Science and Engineering COVID-19 Data Repository. CMS = CARPHA Member States. AIA = Anguilla. ATG = Antigua and Barbuda. BHS = Bahamas. BRB = Barbados. BMU = Bermuda. VGB = British Virgin Islands. CYM = Cayman Islands. DMA = Dominica. GRD = Grenada. GUY = Guyana. JAM = Jamaica. MSR = Montserrat. KNA = Saint Kitts and Nevis. LCA = Saint Lucia. VCT = Saint Vincent and the Grenadines. TTO = Trinidad and Tobago. TCA = Turks and Caicos Islands.

Fig 6. Turnaround times.

(A) The average (across all 17 CMS) percentage time taken from (i) sample collection to receipt by the COVID-19 IMPACT laboratory (red), (ii) laboratory receipt to sequencing (yellow), (iii) sequencing to reporting of results to sending institution (i.e. Trinidad Public Health Laboratory (TPHL) or Caribbean Public Health Agency (CARPHA) (peach), and (iv) from reporting to sending institution to GISAID database upload (light blue) during three periods i.e. 2020, January 1st to June 30th 2021, and July 1st to December 31st 2021. (B) For each country, the average number of days from sample collection to: (i) receipt by the COVID-19 IMPACT laboratory (red circles); (ii) sequencing (yellow circles); (iii) reporting to sending institution (orange circles); and (iv) upload to the GISAID database (light blue circles). The times are shown for samples collected during three periods; 2020, January 1st to June 30th 2021 and July 1st to December 31st 2021. AIA = Anguilla. ATG = Antigua and Barbuda. BHS = Bahamas. BRB = Barbados. BMU = Bermuda. VGB = British Virgin Islands. CYM = Cayman Islands. DMA = Dominica. GRD = Grenada. GUY = Guyana. JAM = Jamaica. MSR = Montserrat. KNA = Saint Kitts and Nevis. LCA = Saint Lucia. VCT = Saint Vincent and the Grenadines. TTO = Trinidad and Tobago. TCA = Turks and Caicos Islands.