Tracking the evolution of biomarker efficacy in SARS-CoV-2: a global meta-analyses series

Background Sophisticated prognostic scores have been proposed for SARS-CoV-2 but do not always perform consistently. We performed this systematic review to discover why and to investigate the impact of vaccination and viral variants.

Methods We searched the PubMed database for the keywords ‘SARS-CoV-2’ or ‘Covid19’ with ‘biomarker’ and ‘mortality’ for the baseline tranche (01/12/2019–30/06/2021) and either ‘SARS-CoV-2’ or ‘Covid19’ with ‘biomarker’ and either ‘vaccination’ or ‘variant’ from 01/12/2020 to 31/10/2023. To aggregate the data, the meta library in R was used, and a random effects model fitted to obtain pooled AUCs and 95% confidence intervals.

Results We screened 4,688 potential source manuscripts and included 144 in the final analyses. Biomarker effectiveness varies significantly by geographical region. Admission CRP levels were a good prognostic marker for mortality due to wild-type virus in Asian countries, with a pooled area under curve (AUC) of 0.83 (95%CI 0.80–0.85), but only an average predictor of mortality in Europe/Northern America, with a pooled AUC of 0.67 (95%CI 0.63–0.71, P < 0.0001). The same pattern applies to D-dimer and IL-6. Notably, urea and troponin had pooled AUCs ≥ 0.78 regardless of location, implying that end-organ damage at presentation was a key prognostic factor in wild-type SARS-CoV-2 infection. CRP, D-dimer, and IL-6 have generally declined in effectiveness in the vaccinated and variant cohorts. We note a significant lag from the pandemic advent to data availability and that the type of data varied considerably between studies.

Conclusion Biomarkers and prognostic scores should be tailored to populations. It is imperative that the infrastructure for collecting clinical data should be put in place ahead of a future pandemic so that data harvesting, collation and analysis can occur in a robust and timely manner to aid clinical decision making.