This study addresses the question of why clinical trial results of anti-SARS-CoV-2 drugs designed to prevent viral replication have been inconsistent. The authors employed a mathematical virus dynamic model describing the viral loads (SARS-CoV-2 RNA copies) over time with differential equations, and determined the model parameters by fitting with real-world viral load data. They found 3 patient groups with significantly different decay rates in the viral load. Next, they simulated the viral load with therapeutic interventions by hypothetical anti-viral-replication agents (50, 95 or 99% inhibition of SARS-CoV-2 replication), at various time points. Their simulation indicated that only treatment at the early onset of viral replication is effective even if the anti-replication agent has 99% efficacy. Finally, they calculated how many patients are needed for clinical trials without considering the viral load decay rate subgroups or the timing of the therapeutic intervention. They concluded that unrealistically large sample sizes are needed under such conditions, and propose that patient populations must be chosen for a reliable evaluation of anti-viral-replication drugs. This study is highly creative. The notable excellence is that they propose a potential solution to overcome the medical problem facing our world today with their excellent expertise in mathematical modeling and publicly available clinical data.
