Viral clearance as a surrogate of clinical efficacy for COVID-19 therapies in outpatients: a systematic review and meta-analysis

Lancet Microbe. 2024 Apr 4:S2666-5247(23)00398-1. doi: 10.1016/S2666-5247(23)00398-1. Online ahead of print.

Abstract

Background: Surrogates of antiviral efficacy are needed for COVID-19. We aimed to investigate the relationship between the virological effect of treatment and clinical efficacy as measured by progression to severe disease in outpatients treated for mild-to-moderate COVID-19.

Methods: In this systematic review and meta-analysis, we searched PubMed, Scopus, and medRxiv from database inception to Aug 16, 2023, for randomised placebo-controlled trials that tested virus-directed treatments (ie, any monoclonal antibodies, convalescent plasma, or antivirals) in non-hospitalised individuals with COVID-19. We only included studies that reported both clinical outcomes (ie, rate of disease progression to hospitalisation or death) and virological outcomes (ie, viral load within the first 7 days of treatment). We extracted summary data from eligible reports, with discrepancies resolved through discussion. We used an established meta-regression model with random effects to assess the association between clinical efficacy and virological treatment effect, and calculated I2 to quantify residual study heterogeneity.

Findings: We identified 1718 unique studies, of which 22 (with a total of 16 684 participants) met the inclusion criteria, and were in primarily unvaccinated individuals. Risk of bias was assessed as low in 19 of 22 studies for clinical outcomes, whereas for virological outcomes, a high risk of bias was assessed in 11 studies, some risk in ten studies, and a low risk in one study. The unadjusted relative risk of disease progression for each extra log10 copies per mL reduction in viral load in treated compared with placebo groups was 0·12 (95% CI 0·04-0·34; p<0·0001) on day 3, 0·20 (0·08-0·50; p=0·0006) on day 5, and 0·53 (0·30-0·94; p=0·030) on day 7. The residual heterogeneity in our meta-regression was estimated as low (I2=0% [0-53] on day 3, 0% [0-71] on day 5, and 0% [0-43] on day 7).

Interpretation: Despite the aggregation of studies with differing designs, and evidence of risk of bias in some virological outcomes, this review provides evidence that treatment-induced acceleration of viral clearance within the first 5 days after treatment is a potential surrogate of clinical efficacy to prevent hospitalisation with COVID-19. This work supports the use of viral clearance as an early phase clinical trial endpoint of therapeutic efficacy.

Funding: Australian Government Department of Health, Medical Research Future Fund, and Australian National Health and Medical Research Council.