Safety, tolerability, viral kinetics, and immune correlates of protection in healthy, seropositive UK adults inoculated with SARS-CoV-2: a single-centre, open-label, phase 1 controlled human infection study

Abstract

Background: A SARS-CoV-2 controlled human infection model (CHIM) has been successfully established in seronegative individuals using a dose of 1×10¹ 50% tissue culture infectious dose (TCID₅₀) pre-alpha SARS-CoV-2 virus. Given the increasing prevalence of seropositivity to SARS-CoV-2, a CHIM that could be used for vaccine development will need to induce infection in those with pre-existing immunity. Our aim was to find a dose of pre-alpha SARS-CoV-2 virus that induced infection in previously infected individuals.

Methods: Healthy, UK volunteers aged 18-30 years, with proven (quantitative RT-PCR or lateral flow antigen test) previous SARS-CoV-2 infection (with or without vaccination) were inoculated intranasally in a stepwise dose escalation CHIM with either 1×10¹, 1×10², 1×10³, 1×10⁴, or 1×10⁵ TCID₅₀ SARS-CoV-2/human/GBR/484861/2020, the same virus used in the seronegative CHIM. Post-inoculation, volunteers were quarantined in functionally negative pressure rooms (Oxford, UK) for 14 days and until 12-hourly combined oropharyngeal-nasal swabs were negative for viable virus by focus-forming assay. Outpatient follow-up continued for 12 months post-enrolment, with additional visits for those who developed community-acquired SARS-CoV-2 infection. The primary objective was to identify a safe, well tolerated dose that induced infection (defined as two consecutive SARS-CoV-2 positive PCRs starting 24 h after inoculation) in 50% of seropositive volunteers. This study is registered with ClinicalTrials.gov (NCT04864548); enrolment and follow-up to 12 months post-enrolment are complete.

Findings: Recruitment commenced on May 6, 2021, with the last volunteer enrolled into the dose escalation cohort on Nov 24, 2022. 36 volunteers were enrolled, with four to eight volunteers inoculated in each dosing group from 1×10¹ to 1×10⁵ TCID₅₀ SARS-CoV-2. All volunteers have completed quarantine, with follow-up to 12 months complete. Despite dose escalation to 1×10⁵ TCID₅₀, we were unable to induce sustained infection in any volunteers. Five (14%) of 36 volunteers were considered to have transient infection, based on the kinetic of their PCR-positive swabs. Transiently infected volunteers had significantly lower baseline mucosal and systemic SARS-CoV-2-specific antibody titres and significantly lower peripheral IFNγ responses against a CD8⁺ T-cell SARS-CoV-2 peptide pool than uninfected volunteers. 14 (39%) of 36 volunteers subsequently developed breakthrough infection with the omicron variant after discharge from quarantine. Most adverse events reported by volunteers in quarantine were mild, with fatigue (16 [44%]) and stuffy nose (16 [44%]) being the most common. There were no serious adverse events.

Interpretation: Our study demonstrates potent protective immunity induced by homologous vaccination and homologous or heterologous previous SARS-CoV-2 infection. The community breakthrough infections seen with the omicron variant supports the use of newer variants to establish a model with sufficient rate of infection for use in vaccine and therapeutic development.

Funding: Wellcome Trust and Department for Health and Social Care.

Figure 1. Study timeline and sample collection

ECG=electrocardiogram. FFA=focus forming assay. GAD-7=7-item Generalized Anxiety Disorder scale. PHQ-9=9-item Patient Health Questionnaire. PFT=pulmonary function test. qRT-PCR=quantitative RT-PCR. SAM=synthetic absorptive matrix (for sampling of nasal lining fluid). TLCO=transfer capacity of the lung for carbon monoxide. UPSIT=University of Pennsylvania Smell Identification Test. WGS=whole-genome sequencing. *Selected participants as detailed in the appendix (study protocol pp 119–121).

Figure 2. Trial profile

TCID₅₀=50% tissue culture infective dose. *QCovid score derived using QCovid tool (University of Oxford, Oxford, UK), a living risk prediction algorithm which gives individuals a calculated risk of hospitalisation and death from COVID-19 (appendix, study protocol pp 78–79).

Figure 3. Positive PCRs during quarantine

The LLOQ for quantitative RT-PCR (qRT-PCR) was 3 log₁₀ copies per mL, with positive detections less than the LLOQ assigned a value of 1·5 log₁₀ copies per mL and undetectable samples assigned a value of 0 log₁₀ copies per mL. For the focus-forming assay (FFA), the LLOQ was 1·57 focus-forming units (FFU) per mL; viral detection less than the LLOQ was assigned 1 log₁₀ FFU per mL; and undetectable samples were assigned 0 log₁₀ FFU per mL. (A) All swabs taken during the quarantine period. Coloured squares denote detection of SARS-CoV-2 by qRT-PCR with viral load values presented; positive detections below the LLOQ are highlighted in yellow with no associated value for viral load. The five volunteers considered to demonstrate transient infection are labelled. All other SARS CoV-2-positive swabs were considered to represent residual inoculum, with initial viral detection occurring on day 1 (denoted by thick black line). (B) Viral kinetics of transiently infected volunteers by qRT-PCR. Dotted lines denote the LLOQ and positive detections below this value. Red dots denote corresponding FFA positivity of that sample, with only one volunteer (002) demonstrating positivity by FFA at a single timepoint (1·82 log₁₀ FFU per mL). LLOQ=lower limit of quantification.

Figure 4. Cellular immunity before and after SARS-CoV-2 infection

(A, B) Ex-vivo SARS CoV-2 peptide-specific PBMC IFNγ ELISpot responses by antigen, comparing transiently infected volunteers with uninfected volunteers. Dots represent individuals, colour coded for infection dose; lines show median with IQR. Background subtracted antigen-specific responses are presented as SFC per million PBMC. (A) Baseline ELISpot responses. (B) AUC analysis of ELISpot response over time between baseline and day 28. Dots represent AUC for individual volunteers. Groups were compared using two sided Mann-Whitney and no significant difference was seen. (C) Ex-vivo SARS CoV-2 peptide-specific ELISpot IFNγ responses by antigen performed on freshly isolated PBMC from volunteers with community-acquired COVID-19 infection during the post-quarantine follow-up period. Dots represent each volunteer; volunteer 017 had two community-acquired infections. Background subtracted antigen-specific responses are presented as SFC per million PBMC. AUC=area under the curve. CD4=CD4⁺ peptide pool. CD8=CD8⁺ peptide pool. M=membrane protein. NP=nucleocapsid protein. ORF=open reading frame. PBMC=peripheral blood mononuclear cells. S1=spike protein subunit 1. S2=spike protein subunit 2. SFC=spot-forming cells. TI=transiently infected. U=uninfected.

Figure 5. Baseline humoral responses against SARS-CoV-2

Baseline antibody responses (IgG, IgA, and IgM) against four SARS-CoV-2 structural proteins, from serum and NLF, comparing transiently infected volunteers with no transient infection. Dots represent individuals, colour coded for infection dose; black lines show median with IQR; red lines denote the pre-pandemic negative cutoff. Background subtracted antigen-specific responses are presented as SFC per million PBMC. AU=arbitrary units. NLF=nasal lining fluid. NP=nucleocapsid protein. NTD=N-terminal domain. RBD=receptor binding domain. S=spike protein. TI=transiently infected. U=uninfected.