Using models to identify the causes of pre-symptomatic transmission from human infection data

When infections can be transmitted from hosts showing no symptoms, containing outbreaks requires distinct strategies like active surveillance. Yet it is rarely clear before-hand when such interventions are needed, especially for emerging pathogens. To investigate the within-host dynamics that enable pre-symptomatic transmission, we survey controlled human infection (CHI) trials with viral pathogens that follow symptoms and viral shedding after inoculation with a known dose. We find that many studies report average timing of symptom onset and shedding, but few report those data for individual participants. We fit a simple model to individual shedding time series from two CHI studies (using norovirus and SARS-CoV-2, respectively) to infer replication rates and the timing of peak shedding relative to symptom onset. We find that faster viral replication significantly hastens peak shedding with minimal impact on symptom onset and no evidence for a tradeoff between the rate and duration of transmission during the pre-symptomatic phase. We then develop and compare within-host models of pathogen replication, immune clearance, and symptom onset to identify plausible assumptions about the causes of pre-symptomatic transmission. We recover the empirical pattern that peak shedding can precede symptom onset when we assume that symptoms are triggered by immune responses rather than pathogen abundance. By incorporating resource limitation via a carrying capacity, we can recover the pattern that faster viral replication prolongs pre-symptomatic transmission. Thus, individual-level data from CHI trials--paired with models--can illuminate the within-host dynamics underpinning pre-symptomatic transmission, guiding efforts to improve control strategies. Significance statementThe COVID-19 pandemic was exacerbated by the potential for transmission before symptoms. Yet the causes of pre-symptomatic transmission--in some hosts but not others--remain unclear, hindering efforts to predict disease spread and tailor control efforts for novel pathogens. To identify patterns across viral taxa, we surveyed con-trolled human infection (CHI) trials, which rarely reported data on the onset of shedding and symptoms for individuals. Individual time series of shedding and symptoms were available from two CHI trials using norovirus and SARS-CoV-2. We fit models to those data to show that faster viral replication hastens shedding but not symptom onset and then used more detailed models to identify plausible assumptions about the within-host causes of underlying pre-symptomatic transmission.