Lowering SARS-CoV-2 viral load might affect transmission but not disease severity in secondary cases - Authors' reply.

We thank Mattia Trunfio and colleagues for their interest in our Personal View regarding the impact of non-pharmaceutical interventions (NPIs) on the viral inoculum of SARS-CoV-2. We agree that increasing evidence supports that NPIs are expected to lower the viral inoculum, potentially contributing to lower transmission. We acknowledge Trunfio and colleagues' point that the evidence supporting the impact of reduced inoculum on COVID-19 severity is less strong than that on infection; we had, therefore, presented this idea as a hypothesis and suggested potential experimental approaches. Of note, human challenge trials have since started in the UK, which will provide more direct evidence on the relationship between viral inoculum and both infection and disease.

We agree that the young age of the participants in Bielecki and colleagues' study is a limitation, although it is not clear how non-airborne routes of transmission would bias the results. The study by Marks and colleagues supports the importance of the index viral load, regardless of symptom status, in forward transmission risk. Although Marks and colleagues did not find a statistically significant association between the index cases' viral loads and the first positive viral loads of the secondary cases (p=0·10), the timing of presentation for symptoms influenced the timing of measurement. Temporal, longitudinal dynamics of PCR cycle thresholds should be accounted for in this type of analysis, given the potential for cycle thresholds to peak before symptoms. Moreover, shedding of viral fragments might not reflect the true inoculum, with additional viral culture studies needed.

We disagree that the referenced challenge study in rhesus macaques provides conflicting results on the dose–response relationship. A single dosage (nCoV-WA1-2020 isolate) was provided in this animal study and was not systematically varied in a controlled manner. Therefore, information on the dose–response relationship cannot be inferred from this study. In our Personal View, we suggest experimental approaches in animal models that could explore this hypothesis further—ie, systematically varying the inoculum dose, confirming successful infection using viral culture or molecular methods, and then presenting data on clinical outcomes among animals that were successfully infected.

We agree that host factors such as age and chronic medical conditions are key factors in SARS-CoV-2 susceptibility. However, as these factors are generally not modifiable, we argue that further research is needed to explore the relationship between NPIs and the viral inoculum. Such exploration could provide additional evidence supporting the use of NPIs in COVID-19 mitigation. Given the need to protect unvaccinated individuals and reduce transmission while vaccination distribution continues, this research hypothesis merits continued focus.

MAS and MG are funded by NIAID/NIH R01AI158013. We declare no competing interests.