Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


Understanding COVID-19 through human challenge models

As the pandemic evolves, questions remain but field studies are increasingly difficult; if carefully designed, human challenge models could offer a practical and timely solution.

Controlled human infection models (CHIMs) have been used to study the pathogenesis of disease and evaluate preventive and therapeutic approaches for many years1. From the onset of the COVID-19 pandemic, vaccinologists, ethicists and clinical trial researchers have debated the risks and merits of this approach2,3,4,5, including a World Health Organization (WHO) task force that convened to outline the potential ethical and operational issues6,7. The central tenet of any CHIM involves balancing the scientific and public health benefits that might accrue with any adverse events that could develop in study participants — a particular concern with the marked morbidity and mortality associated with SARS-CoV-2 infection. In this issue of Nature Medicine, Killingley et al.8 present results from the first human challenge study with SARS-CoV-2.

In the study, healthy adults between the ages of 18 and 29 years — without serological evidence of previous SARS-CoV-2 infection or vaccination — were intranasally infected with a low dose inoculum of a wild-type SARS CoV-2 strain8. In total, 18 of the 34 seronegative volunteers (53%) became infected, with the viral load peaking at 5 days after inoculation. Although SARS-CoV-2 was initially detected in the throat, it subsequently rose to higher titers in the nose and persisted for 10 days. Mild to moderate respiratory symptoms with anosmia (loss of sense of smell) occurred in most of the infected participants, but no clinical or radiological evidence of pulmonary involvement was documented. The first cohort of six challenged participants received the antiviral drug remdesivir, but as symptoms and viral loads did not differ with treatment, subsequent volunteers were untreated. At 180 days after inoculation, 5 out of 12 participants with anosmia had persistent but improved symptoms. No other long-term adverse events were reported.

The manufacturing of challenge viruses and the development of CHIMs typically takes years, and thus the speed of this large and complicated endeavor is remarkable. The steps to study conduct included the preparation of a Good Manufacturing Practice (GMP)-compliant challenge stock; meticulous attention to the preparation and administration of the inoculum; careful participant selection with comprehensive informed consent and assurance that multispecialty care would be provided if adverse events occurred; strict participant isolation procedures; and rigid adherence to regulatory guidelines9 throughout. Comprehensive immunological and virological studies were also included in the study design (although these data are not yet available).

The current report largely focused on SARS-CoV-2 viral kinetics, early neutralizing antibody responses to infection and utility of lateral flow assay rapid antigen tests to detect early asymptomatic infection and reduce transmission. These careful analyses also documented the safety of this approach in a small number of highly selected participants. However, recently reported data from a prospective cohort-based imaging study detected a subtle but consistent spatial pattern of abnormalities in regions of the brain associated with the olfactory network in participants with confirmed COVID-19 infections10 — therefore, short- and long-term assessment of abnormalities in brain imaging or function should be considered in those participating in the CHIM studies. Despite this caution, the current study does provide a usable model for future research. However, many of the most intriguing questions remain to be addressed — for example, what are the antibody and cellular immune correlates of protection against infection? What is the mechanism behind some seronegative participants being resistant to challenge infection? Frequent sampling is a clear advantage of challenge studies over field studies, and archival samples from the current study will be used to dissect the kinetics of local and systemic immune responses, to try and answer some of these questions.

But can these models be extended to address additional burning questions relating to SARS-CoV-2? The evolution of the pandemic has created a need for alternatives to field studies. Early in the pandemic, with a highly susceptible population and a high viral attack rate, prospective studies and clinical trials were able to provide rapid information on the clinical course of illness and the efficacy of vaccines and treatments. The early vaccine candidates were evaluated in this way, with trials including adults of all ages, with and without medical co-morbidities, from diverse racial and ethnic backgrounds. However, as the pandemic progresses, vaccine evaluation is becoming increasingly complicated. In contrast to the first available vaccines, dozens of the newer candidates in early stages of development now face uncertainty; extensive public health measures to reduce contagion and increasing vaccination coverage mean fewer immunologically naive population groups available for study. Thus, the feasibility of timely field efficacy trials has come under question, and placebo-controlled trials are increasingly difficult to justify.

Many new vaccines, monoclonal antibodies and antivirals have not yet undergone clinical trials and remain investigational. Although the CHIM may represent an attractive solution to the problems faced in current field studies, this model also has its limitations. In contrast to field trials, challenge studies are generally limited to the healthy, younger population, making generalizability to other populations less reliable. Also, if the CHIM were to be used to assess each emerging variant of concern, the GMP production of such strains would be expensive and potentially slow; this has been the experience with influenza challenge models, making their timely use to address vaccine questions far from optimal11.

Nonetheless, CHIMs could be used to initially evaluate vaccines and prioritize those that seem the most promising. As the number of vaccinated people increases, CHIMs could be used to assess the duration of vaccine protection, and as antigenically diverse variants emerge, cross-strain protection could be evaluated in this model. Individuals with pre-existing immunity could also be challenged with escalating doses of the challenge strain or re-challenged with new strains to understand cross-protective immunity. The CHIM model could be extended to address even more research questions. Transmission studies could be extended to correlate viral detection by various methods (PCR, culture) with transmission; levels of virus generated by aerosols from infected individuals could also be determined. In the context of antivirals, monoclonal antibodies and other therapeutic agents, challenge models could refine the optimal dose and assess the effect of the therapy on symptomatic and asymptomatic infection, and on symptom severity and duration. In addition, immunological parameters could be further dissected and correlates of antibody or T cell protection determined; new diagnostic methods could also be evaluated and compared with existing ones.

Given the many questions that remain, and the devastating toll of this pandemic, we endorse the development and refinement of human challenge models for SARS-CoV-2 and other coronaviruses as part of the pandemic response and preparedness efforts2. We eagerly await the additional planned analyses with samples collected in this trial to dissect “host factors present at the time of inoculation and associated with protection” from infection8. Killingley et al.8 also state that further challenge studies are underway in both vaccinated and previously infected individuals to explore factors that influence infection and clinical outcome. It is hoped that this model can be productively used to speed up the evaluation of vaccines, therapeutic agents and diagnostics and their rapid deployment to the community.


  1. Darton, T. C. et al. Lancet Infect. Dis. 15, 840–851 (2015).

    Article  Google Scholar 

  2. Deming, M. E., Michael, N. L., Robb, M., Cohen, M. S. & Neuzil, K. M. N. Engl. J. Med 383, e63 (2020).

    CAS  Article  Google Scholar 

  3. Eyal, N., Lipsitch, M. & Smith, P. G. J. Infect. Dis. 21, 1752–1756 (2020).

    Article  Google Scholar 

  4. Shah, S. K. et al. Science 368, 832–834 (2020).

    CAS  Article  Google Scholar 

  5. Plotkin, S. A. & Caplan, A. Vaccine 38, 3987–3988 (2020).

    CAS  Article  Google Scholar 

  6. Levine, M. M. et al. Clin. Infect. Dis. 72, 2035–2041 (2021).

    CAS  Article  Google Scholar 

  7. WHO. (6 May 2020).

  8. Killingley, B. et al. Nat. Med. (2022).

    Article  PubMed  Google Scholar 

  9. Rapeport, G. et al. N. Engl. J. Med 385, 961–964 (2021).

    CAS  Article  Google Scholar 

  10. Douaud, G. et al. Nature (2022).

    Article  PubMed  PubMed Central  Google Scholar 

  11. Memoli, M. J. et al. Clin. Infect. Dis. 60, 693–702 (2015).

    CAS  Article  Google Scholar 

Download references

Author information

Authors and Affiliations


Corresponding authors

Correspondence to Kathryn M. Edwards or Kathleen M. Neuzil.

Ethics declarations

Competing interests

K.M.E. was a member of the Data Safety and Monitoring Board for the Pfizer-BioNTech COVID-19 mRNA vaccine and received funding for this work. Her institution received support from the Centers for Disease Control to assess safety of the COVID 19 vaccines and she received salary support for these efforts. K.M.N. was an investigator and her institution received funding for the phase 1 trial of the Pfizer-BioNTech COVID-19 mRNA vaccine. She received no salary support. K.M.N. has received salary support from the National Institutes of Health for her role in the design and execution of the COVID-19 trials funded as part of the COVID-19 Prevention Network.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Edwards, K.M., Neuzil, K.M. Understanding COVID-19 through human challenge models. Nat Med 28, 903–904 (2022).

Download citation

  • Published:

  • Issue Date:

  • DOI:


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing