Diagnostic performance of allele-specific RT-qPCR and genomic sequencing in wastewater-based surveillance of SARS-CoV-2
Md Pervez Kabir, Élisabeth Mercier, Walaa Eid, Julio Plaza‐Díaz, Patrick M. D’Aoust, Chrystal Landgraff, Lawrence Goodridge, Opeyemi U. Lawal, Shen Wan, Nada Hegazy, Tram Nguyen, Chandler Wong, Ocean Thakali, Lakshmi Pisharody, Sean Stephenson, Tyson E. Graber, Robert Delatolla
Abstract
Clinical genomic surveillance is regarded as the gold standard for monitoring SARS-CoV-2 variants globally. However, as the pandemic wanes, reduced testing poses a risk to effectively tracking the trajectory of these variants within populations. Wastewater-based genomic surveillance that estimates variant frequency based on its defining set of alleles derived from clinical genomic surveillance has been successfully implemented. This method has its challenges, and allele-specific (AS) RT-qPCR or RT-dPCR may instead be used as a complementary method for estimating variant prevalence. Demonstrating equivalent performance of these methods is a prerequisite for their continued application in current and future pandemics. Here, we compared single-allele frequency using AS-RT-qPCR, to single-allele or haplotype frequency estimations derived from amplicon-based sequencing to estimate variant prevalence in wastewater during emergent and prevalent periods of Delta, Omicron, and two sub-lineages of Omicron. We found that all three methods of frequency estimation were concordant and contained sufficient information to describe the trajectory of variant prevalence. We further confirmed the accuracy of these methods by quantifying the diagnostic performance through Youden's index. The Youden's index of AS-RT-qPCR was reduced during the low prevalence period of a particular variant while the same allele in sequencing was negatively influenced due to insufficient read depth. Youden's index of haplotype-based calls was negatively influenced when alleles were common between variants. Coupling AS-RT-qPCR with sequencing can overcome the shortcomings of either platform and provide a comprehensive picture to the stakeholders for public health responses. • SARS-CoV-2 variant frequency estimation in wastewater is comparable between AS-RT-qPCR and genomic sequencing. • Diagnostic performance confirms the accuracy of AS-RT-qPCR and genomic sequencing for variant estimation in wastewater. • Diagnostic performance of AS-RT-qPCR is reduced during periods of lower SARS-CoV-2 variant concentration in wastewater. • Diagnostic performance of wastewater sequencing is reduced due to insufficient sequencing depth of the targeted allele. • Monitoring of more than one allele is suggested to overcome limitations of both AS-RT-qPCR and sequencing. • Accuracy of variant frequency estimation by sequenced haplotype requires careful curation.