Ninety-day outcome of patients with severe COVID-19 treated with tocilizumab - a single centre cohort study.

OBJECTIVES: Patients with severe COVID-19 may be at risk of longer term sequelae. Long-term clinical, immunological, pulmonary and radiological outcomes of patients treated with anti-inflammatory drugs are lacking. METHODS: In this single-centre prospective cohort study, we assessed 90-day clinical, immunological, pulmonary and radiological outcomes of hospitalised patients with severe COVID-19 treated with tocilizumab from March 2020 to May 2020. Criteria for tocilizumab administration were oxygen saturation <93%, respiratory rate >30/min, C-reactive protein levels >75 mg/l, extensive area of ground-glass opacities or progression on computed tomography (CT). Descriptive analyses were performed using StataIC 16. RESULTS: Between March 2020 and May 2020, 50 (27%) of 186 hospitalised patients had severe COVID-19 and were treated with tocilizumab. Of these, 52% were hospitalised on the intensive care unit (ICU) and 12% died. Eleven (22%) patients developed at least one microbiologically confirmed super-infection, of which 91% occurred on ICU. Median duration of hospitalisation was 15 days (interquartile range [IQR] 10–24) with 24 days (IQR 14–32) in ICU patients and 10 days (IQR 7–15) in non-ICU patients. At day 90, 41 of 44 survivors (93%) were outpatients. No long-term adverse events or late-onset infections were identified after acute hospital care. High SARS-CoV-2 antibody titres were found in all but one patient, who was pretreated with rituximab. Pulmonary function tests showed no obstructive patterns, but restrictive patterns in two (5.7%) and impaired diffusion capacities for carbon monoxide in 11 (31%) of 35 patients, which predominated in prior ICU patients. Twenty-one of 35 (60%) CT-scans at day 90 showed residual abnormalities, with similar distributions between prior ICU and non-ICU patients. CONCLUSIONS: In this cohort of severe COVID-19 patients, no tocilizumab-related long-term adverse events or late-onset infections were identified. Although chest CT abnormalities were highly prevalent at day 90, the majority of patients showed normal lung function. TRIAL REGISTRATION: ClinicalTrials.gov NCT04351503.

Optimized intramuscular immunization with VSV-vectored spike protein triggers a superior immune response to SARS-CoV-2.

Immunization with vesicular stomatitis virus (VSV)-vectored COVID-19 vaccine candidates expressing the SARS-CoV-2 spike protein in place of the VSV glycoprotein relies implicitly on expression of the ACE2 receptor at the muscular injection site. Here, we report that such a viral vector vaccine did not induce protective immunity following intramuscular immunization of K18-hACE2 transgenic mice. However, when the viral vector was trans-complemented with the VSV glycoprotein, intramuscular immunization resulted in high titers of spike-specific neutralizing antibodies. The vaccinated animals were fully protected following infection with a lethal dose of SARS-CoV-2-SD614G via the nasal route, and partially protected if challenged with the SARS-CoV-2Delta variant. While dissemination of the challenge virus to the brain was completely inhibited, replication in the lung with consequent lung pathology was not entirely controlled. Thus, intramuscular immunization was clearly enhanced by trans-complementation of the VSV-vectored vaccines by the VSV glycoprotein and led to protection from COVID-19, although not achieving sterilizing immunity.

Potent neutralization by monoclonal human IgM against SARS-CoV-2 is impaired by class switch.

To investigate the class-dependent properties of anti-viral IgM antibodies, we use membrane antigen capture activated cell sorting to isolate spike-protein-specific B cells from donors recently infected with SARS-CoV-2, allowing production of recombinant antibodies. We isolate 20, spike-protein-specific antibodies of classes IgM, IgG, and IgA, none of which shows any antigen-independent binding to human cells. Two antibodies of class IgM mediate virus neutralization at picomolar concentrations, but this potency is lost following artificial switch to IgG. Although, as expected, the IgG versions of the antibodies appear to have lower avidity than their IgM parents, this is not sufficient to explain the loss of potency.