Epidemiological and virological factors determining dengue transmission in Sri Lanka during the COVID-19 pandemic

With the onset of the COVID-19 pandemic in early 2020 there was a drastic reduction in the number of dengue cases in Sri Lanka, with an increase towards the end of 2021. We sought to study the contribution of virological factors, human mobility, school closure and mosquito factors in affecting these changes in dengue transmission in Sri Lanka during this time. To understand the reasons for the differences in the dengue case numbers in 2020 to 2021 compared to previous years, we determined the association between the case numbers in Colombo (which has continuously reported the highest number of cases) with school closures, stringency index, changes in dengue virus (DENV) serotypes and vector densities. There was a 79.4% drop in dengue cases from 2019 to 2020 in Colombo. A significant negative correlation was seen with the number of cases and school closures (Spearman's r = -0.4732, p <0.0001) and a negative correlation, which was not significant, between the stringency index and case numbers (Spearman's r = -0.3755 p = 0.0587). There was no change in the circulating DENV serotypes with DENV2 remaining the most prevalent serotype by early 2022 (65%), similar to the frequencies observed by end of 2019. The Aedes aegypti premise and container indices showed positive but insignificant correlations with dengue case numbers (Spearman r = 0.8827, p = 0.93). Lockdown measures, especially school closures seemed to have had a significant impact on the number of dengue cases, while the vector indices had a limited effect.

Epidemiological and virological factors determining dengue transmission in Sri Lanka during the COVID-19 pandemic.

With the onset of the COVID-19 pandemic in early 2020 there was a drastic reduction in the number of dengue cases in Sri Lanka, with an increase towards the end of 2021. We sought to study the contribution of virological factors, human mobility, school closure and mosquito factors in affecting these changes in dengue transmission in Sri Lanka during this time. To understand the reasons for the differences in the dengue case numbers in 2020 to 2021 compared to previous years, we determined the association between the case numbers in Colombo (which has continuously reported the highest number of cases) with school closures, stringency index, changes in dengue virus (DENV) serotypes and vector densities. There was a 79.4% drop in dengue cases from 2019 to 2020 in Colombo. A significant negative correlation was seen with the number of cases and school closures (Spearman's r = -0.4732, p <0.0001) and a negative correlation, which was not significant, between the stringency index and case numbers (Spearman's r = -0.3755 p = 0.0587). There was no change in the circulating DENV serotypes with DENV2 remaining the most prevalent serotype by early 2022 (65%), similar to the frequencies observed by end of 2019. The Aedes aegypti premise and container indices showed positive but insignificant correlations with dengue case numbers (Spearman r = 0.8827, p = 0.93). Lockdown measures, especially school closures seemed to have had a significant impact on the number of dengue cases, while the vector indices had a limited effect.

Efficacy of rupatadine in reducing the incidence of dengue haemorrhagic fever in patients with acute dengue: A randomised, double blind, placebo-controlled trial.

BACKGROUND: Rupatadine was previously shown to reduce endothelial dysfunction in vitro, reduced vascular leak in dengue mouse models and to reduce the extent of pleural effusions and thrombocytopenia in patients with acute dengue. Therefore, we sought to determine the efficacy of rupatadine in reducing the incidence of dengue haemorrhagic fever (DHF) in patients with acute dengue. METHODS AND FINDINGS: A phase 2, randomised, double blind, placebo controlled clinical trial was carried out in patients with acute dengue in Sri Lanka in an outpatient setting. Patients with ≤3 days since the onset of illness were either recruited to the treatment arm of oral rupatadine 40mg for 5 days (n = 123) or the placebo arm (n = 126). Clinical and laboratory features were measured daily to assess development of DHF and other complications. 12 (9.7%) patients developed DHF in the treatment arm compared to 22 (17.5%) who were on the placebo although this was not significant (p = 0.09, relative risk 0.68, 95% CI 0.41 to 1.08). Rupatadine also significantly reduced (p = 0.01) the proportion of patients with platelet counts <50,000 cells/mm3 and significantly reduced (p = 0.04) persisting vomiting, headache and hepatic tenderness (p<0.0001) in patients. There was a significant difference in the duration of illness (p = 0.0002) although the proportion of individuals who required hospital admission in both treatment arms. Only 2 patients on rupatadine and 3 patients on the placebo developed shock, while bleeding manifestations were seen in 6 patients on rupatadine and 7 patients on the placebo. CONCLUSIONS: Rupatadine appeared to be safe and well tolerated and showed a trend towards a reducing proportion of patients with acute dengue who developed DHF. Its usefulness when used in combination with other treatment modalities should be explored. TRIAL REGISTRATION: International Clinical Trials Registration Platform: SLCTR/2017/024.

Comparison of the immunogenicity of five COVID-19 vaccines in Sri Lanka.

To determine the antibody responses elicited by different vaccines against SARS-CoV-2, we compared antibody responses in individuals 3 months post-vaccination in those who had received different vaccines in Sri Lanka. Abs to the receptor binding domain (RBD) of the ancestral (wild type) virus (WT) as well as to variants of concern (VoCs), and ACE2 blocking Abs, were assessed in individuals vaccinated with Moderna (n = 225), Sputnik V (n = 128) or Sputnik light (n = 184) and the results were compared with previously reported data on Sinopharm and AZD1222 vaccinees. A total of 99.5% of Moderna, >94% of AZD1222 or Sputnik V and >70% of Sputnik light, >60% of Sinopharm vaccine recipients, had a positive response to ACE2 blocking antibodies. The ACE2 blocking antibody levels were highest to lowest was Moderna > Sputnik V/AZD1222 (had equal levels) > Sputnik light > Sinopharm. All Moderna recipients had antibodies to the RBD of WT, alpha and beta, while positivity rates for delta variant was 80%. The positivity rates for Sputnik V vaccinees for the WT and VoCs were higher than for AZD1222 vaccinees while those who received Sinopharm had the lowest positivity rates (<16.7%). The total antibodies to the RBD were highest for the Sputnik V and AZD1222 vaccinees. The Moderna vaccine elicited the highest ACE2 blocking antibody levels followed by Sputnik V/AZD1222, while those who received Sinopharm had the lowest levels. These findings highlight the need for further studies to understand the effects on clinical outcomes.

Comparison of different sequencing techniques for identification of SARS-CoV-2 variants of concern with multiplex real-time PCR.

As different SARS-CoV-2 variants emerge and with the continuous evolvement of sub lineages of the delta variant, it is crucial that all countries carry out sequencing of at least >1% of their infections, in order to detect emergence of variants with higher transmissibility and with ability to evade immunity. However, due to limited resources as many resource poor countries are unable to sequence adequate number of viruses, we compared to usefulness of a two-step commercially available multiplex real-time PCR assay to detect important single nucleotide polymorphisms (SNPs) associated with the variants and compared the sensitivity, accuracy and cost effectiveness of the Illumina sequencing platform and the Oxford Nanopore Technologies' (ONT) platform. 138/143 (96.5%) identified as the alpha and 36/39 (92.3%) samples identified as the delta variants due to the presence of lineage defining SNPs by the multiplex real time PCR, were assigned to the same lineage by either of the two sequencing platforms. 34/37 of the samples sequenced by ONT had <5% ambiguous bases, while 21/37 samples sequenced using Illumina generated <5%. However, the mean PHRED scores averaged at 32.35 by Illumina reads but 10.78 in ONT. This difference results in a base error probability of 1 in 10 by the ONT and 1 in 1000 for Illumina sequencing platform. Sub-consensus single nucleotide variations (SNV) are highly correlated between both platforms (R2 = 0.79) while indels appear to have a weaker correlation (R2 = 0.13). Although the ONT had a slightly higher error rate compared to the Illumina technology, it achieved higher coverage with a lower number or reads, generated less ambiguous bases and was significantly less expensive than Illumina sequencing technology.

Kinetics of immune responses to the AZD1222/Covishield vaccine with varying dose intervals in Sri Lankan individuals.

BACKGROUND: To understand the kinetics of immune responses with different dosing gaps of the AZD1222 vaccine, we compared antibody and T cell responses in two cohorts with two different dosing gaps. METHODS: Antibodies to the SARS-CoV-2 virus were assessed in 297 individuals with a dosing gap of 12 weeks, sampled 12 weeks post second dose (cohort 1) and in 77 individuals with a median dosing gap of 21.4 weeks (cohort 2) sampled 6 weeks post second dose. ACE2-blocking antibodies (ACE2-blocking Abs), antibodies to the receptor-binding domain (RBD) of  variants of concern (VOC), and ex vivo T cell responses were assessed in a subcohort. RESULTS: All individuals (100%) had SARS-CoV-2-specific total antibodies and 94.2% of cohort 1 and 97.1% of cohort 2 had ACE2-blocking Abs. There was no difference in antibody titers or positivity rates in different age groups in both cohorts. The ACE2-blocking Abs (p < .0001) and antibodies to the RBD of the VOCs were significantly higher in cohort 2 compared to cohort 1. 41.2% to 65.8% of different age groups gave a positive response by the hemagglutination assay to the RBD of the ancestral virus and VOCs in cohort 1, while 53.6%-90% gave a positive response in cohort 2. 17/57 (29.8%) of cohort 1 and 17/29 (58.6%) of cohort 2 had ex vivo interferon (IFN)γ ELISpot responses above the positive threshold. The ACE2-blocking antibodies (Spearman's r = .46, p = .008) and ex vivo IFNγ responses (Spearman's r = .71, p < .0001) at 12 weeks post first dose, significantly correlated with levels 12 weeks post second dose. CONCLUSIONS: Both dosing schedules resulted in high antibody and T cell responses post vaccination, although those with a longer dosing gap had a higher magnitude of responses, possibly as immune responses were measured 6 weeks post second dose compared to 12 weeks post second dose.

Sensitivity and specificity of two WHO approved SARS-CoV2 antigen assays in detecting patients with SARS-CoV2 infection.

BACKGROUND: SARS-CoV-2 rapid antigen (Ag) detection kits are widely used in addition to quantitative reverse transcription PCR PCR (RT-qPCR), as they are cheaper with a rapid turnaround time. As there are many concerns regarding their sensitivity and specificity, in different settings, we evaluated two WHO approved rapid Ag kits in a large cohort of Sri Lankan individuals. METHODS: Paired nasopharangeal swabs were obtained from 4786 participants for validation of the SD-Biosensor rapid Ag assay and 3325 for the Abbott rapid Ag assay, in comparison to RT-qPCR. A short questionnaire was used to record symptoms at the time of testing, and blood samples were obtained from 2721 of them for detection of SARS-CoV-2 specific antibodies. RESULTS: The overall sensitivity of the SD-Biosensor Ag kit was 36.5% and the Abbott Ag test was 50.76%. The Abbott Ag test showed specificity of 99.4% and the SD-Biosensor Ag test 97.5%. At Ct values < 25, the sensitivity was 71.3% to 76.6% for the SD-Biosensor Ag test and 77.3% to 88.9% for the Abbott Ag test. The Ct values for all genes (RdRP, S, E and N) tested with all RT-qPCR kits were significantly lower for the positive results of the Abbott Ag test compared to the SD-Biosensor test. 209 (48.04%) individuals who had antibodies gave a positive RT-qPCR result, and antibody positivity rates were higher at Ct values > 30 (46.1 to 82.9%). 32.1% of those who gave a positive result with the SD-Biosensor Ag test and 26.3% of those who gave positive results with the Abbott Ag test had SARS-CoV-2 antibodies at the time of detection. CONCLUSIONS: Both rapid Ag tests appeared to be highly sensitive in detecting individuals at lower Ct values, in a community setting in Sri Lanka, but it will be important to further establish the relationship to infectivity.

Kinetics of immune responses to SARS-CoV-2 proteins in individuals with varying severity of infection and following a single dose of the AZD1222.

To characterize the IgG and IgA responses to different SARS-CoV-2 proteins, we investigated the antibody responses to SARS-CoV-2 following natural infection and following a single dose of AZD1222 (Covishield), in Sri Lankan individuals. The IgG and IgA responses were assessed to S1, S2, RBD, and N proteins in patients at 4 weeks and 12 weeks since the onset of illness or following vaccination. Antibodies to the receptor-binding domain of SARS-CoV-2 wild type (WT), α, ß, and λ and ACE2 (Angiotensin Converting Enzyme 2) receptor blocking antibodies were also assessed in these cohorts. For those with mild illness and in vaccines, the IgG responses to S1, S2, RBD, and N protein increased from 4 weeks to 12 weeks, while it remained unchanged in those with moderate/severe illness. In the vaccines, IgG antibodies to the S2 subunit had the highest significant rise (P < 0.0001). Vaccines had several-fold lower IgA antibodies to all the SARS-CoV-2 proteins tested than those with natural infection. At 12 weeks, the haemagglutination test (HAT) titres were significantly lower to the α in vaccines and significantly lower in those with mild illness and in vaccines to ß and for λ. No such difference was seen in those with moderate/severe illness. Vaccines had significantly less IgA to SARS-CoV-2, but comparable IgG responses those with natural infection. However, following a single dose vaccines had reduced antibody levels to the VOCs, which further declined with time, suggesting the need to reduce the gap between the two doses, in countries experiencing outbreaks due to VOCs.

Kinetics of immune responses to SARS-CoV-2 proteins in individuals with varying severity of infection and following a single dose of the AZD1222

To characterize the IgG and IgA responses to different SARS-CoV-2 proteins, we investigated the antibody responses to SARS-CoV-2 following natural infection and following a single dose of AZD1222(Covishield), in Sri Lankan individuals. The IgG and IgA responses were assessed to S1, S2, RBD and N proteins in patients at 4 weeks and 12 weeks since onset of illness or following vaccination. Antibodies to the receptor binding domain of SARS-CoV-2 wild type (WT), alpha, beta and delta and ACE2 (Angiotensin Converting Enzyme 2) receptor blocking antibodies were also assessed in these cohorts. Those with mild illness and in vaccinees, the IgG responses to S1, S2, RBD and N protein increased from 4 weeks to 12 weeks, while it remained unchanged in those with moderate/severe illness. In the vaccinees, IgG antibodies to the S2 subunit had the highest significant rise(p<0.0001). Vaccinees had several fold lower IgA antibodies to all the SARS-CoV-2 proteins tested than those with natural infection. At 12 weeks, the Haemagglutination test (HAT) titres were significantly lower to the alpha in vaccinees and significantly lower in those with mild illness and in vaccinees to beta and for delta. No such difference was seen in those with moderate/severe illness. Vaccinees had significantly less IgA to SARS-CoV-2, but comparable IgG responses those with natural infection. However, following a single dose vaccinees had reduced antibody levels to the VOCs, which further declined with time, suggesting the need to reduce the gap between the two doses, in countries experiencing outbreaks due to VOCs.

Surveillance of SARS-CoV-2 variants of concern by identification of single nucleotide polymorphisms in the spike protein by a multiplex real-time PCR.

BACKGROUND: Although active surveillance SARS-CoV-2 variants of concern (VOCs) is required for proper outbreak control measures, many lower income countries find it challenging to detect VOCs by carrying genomic sequencing alone, due to limited resources. METHODS: VOCs can also be identified by the unique mutations in the spike protein by real-time PCR that detect these single nucleotide polymorphisms (SNPs). We used a multiplex, real-time PCR assay for detection of these SNPs for identification of the prevalence of different SARS-CoV-2 VOCs in 16/26 districts in Sri Lanka. RESULTS: Of the 664/934 that were subjected to the multiplex qRT-PCR, 638 (96.1 %) detected L452R and K417 in the channels and were identified as the delta variant. 25 samples (3.9 %) detected N501Y, with K417 were considered as the alpha variant. Of 10/16 districts in Sri Lanka, the delta variant was the only VOC detected. CONCLUSIONS: This multiplex real-time qRT-PCR which identifies certain SNPs specific to the VOCs appears to be a fast, cheaper and less technically demanding method to generate data regarding the spread of different SARS-CoV-2 variants, and is a suitable method for lower income countries, to supplement the data generated by genomic sequencing.

Genomic and Epidemiological Analysis of SARS-CoV-2 Viruses in Sri Lanka.

Background: In order to understand the molecular epidemiology of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) in Sri Lanka, since March 2020, we carried out genomic sequencing overlaid on available epidemiological data until April 2021. Methods: Whole genome sequencing was carried out on diagnostic sputum or nasopharyngeal swabs from 373 patients with COVID-19. Molecular clock phylogenetic analysis was undertaken to further explore dominant lineages. Results: The B.1.411 lineage was most prevalent, which was established in Sri Lanka and caused outbreaks throughout the country until March 2021. The estimated time of the most recent common ancestor (tMRCA) of this lineage was June 1, 2020 (with 95% lower and upper bounds March 30 to July 27) suggesting cryptic transmission may have occurred, prior to a large epidemic starting in October 2020. Returning travellers were identified with infections caused by lineage B.1.258, as well as the more transmissible B.1.1.7 lineage, which has replaced B.1.411 to fuel the ongoing large outbreak in the country. Conclusions: The large outbreak that started in early October, is due to spread of a single virus lineage, B.1.411 until the end of March 2021, when B.1.1.7 emerged and became the dominant lineage.

SARS-CoV-2 neutralizing antibodies in patients with varying severity of acute COVID-19 illness.

In order to support vaccine development, and to aid convalescent plasma therapy, it would be important to understand the kinetics, timing and persistence of SARS-CoV-2 neutralizing antibodies (NAbs), and their association with clinical disease severity. Therefore, we used a surrogate viral neutralization test to evaluate their levels in patients with varying severity of illness, in those with prolonged shedding and those with mild/asymptomatic illness at various time points. Patients with severe or moderate COVID-19 illness had earlier appearance of NAbs at higher levels compared to those with mild or asymptomatic illness. Furthermore, those who had prolonged shedding of the virus, had NAbs appearing faster and at higher levels than those who cleared the virus earlier. During the first week of illness the NAb levels of those with mild illness was significantly less (p = 0.01), compared to those with moderate and severe illness. At the end of 4 weeks (28 days), although 89% had NAbs, 38/76 (50%) in those with > 90 days had a negative result for the presence of NAbs. The Ab levels significantly declined during convalescence (> 90 days since onset of illness), compared to 4 to 8 weeks since onset of illness. Our data show that high levels of NAbs during early illness associated with clinical disease severity and that these antibodies declined in 50% of individuals after 3 months since onset of illness.

Similarities and differences between the 'cytokine storms' in acute dengue and COVID-19.

Severe pneumonia and multiorgan dysfunction in COVID-19 and dengue haemorrhagic fever (DHF) are two diseases that can associate with an altered immune response to the infecting virus. To determine the similarities and differences in the cytokine and chemokine responses in these two infections, we compared responses in patients with varying severity of COVID-19 and acute dengue at different time points of illness. During early disease, patients who proceeded to develop COVID-19 severe pneumonia (SP) and DHF had significantly higher levels of IL-6, IL-10 and MIP3α than those who developed mild illness. The lowest levels of IFNγ in early illness were seen in those who succumbed to their illness due to COVID-19. Levels of serum IL-10 (p = 0.0001), IL-6 (p = 0.002), MIP-3α (p = 0.02) and CD40-L levels (p = 0.002) significantly increased from 5 to 9 day of illness to 10-21 day of illness in patients with moderate-to-severe COVID-19, but not in those with mild illness. In contrast, these cytokine/chemokine levels remained unchanged in those with DHF or dengue fever (DF) during febrile and critical phases. Although IL-10 levels were significantly higher in COVID-19 patients with SP, patients with DHF had 25-fold higher levels, whereas IL-6 levels were 11-fold higher in those with COVID-19 SP. IL-10 and other cytokines were evaluated in a larger cohort of patients during early illness (≤ 4 days) who proceeded to develop DF (n = 71) or DHF (n = 64). Of the cytokines evaluated, IL-10 was significantly higher (p < 0.0001) in those who went on to develop DHF compared to DF. Low IFNγ response to the SARS-CoV2 and high levels of immunosuppressive IL-10 in both COVID-19 and dengue during early illness are indicators of an altered antiviral response potentially contributing to disease severity.