LAUSR

Convalescent plasma has been a mainstay therapeutic in passive immunization for decades. In the setting of the novel coronavirus (SARS-CoV-2), it has become widely used for prophylaxis and early intervention in COVID-19 [1, 2]. Recently, neutralizing autoantibodies to type I interferons have been described in at least 10% of patients with critical COVID-19 pneumonia, while they were absent from infected individuals with asymptomatic or mild disease [3]. These autoantibodies are likely pre-existing and have an immunological impact early in the course of COVID-19 [3,4]. Given the role of auto-Abs to type I interferons in the development of life-threatening COVID-19 pneumonia, we determined the prevalence of anti-type I interferon antibodies in the convalescent plasma supply from a large blood bank. We selected 116 convalescent plasma samples from unique donors who had previously been hospitalized for COVID-19, and found that 4/116 (3%) plasma samples were positive for anti-IFN-α2 autoantibodies (Fig. 1a). All 4 positive patients were male and 50–70 years of age (Supplemental Table 1). We further tested these 4 samples for their neutralizing capacity against 10 ng/ml of IFN-α2 and IFN-ω in a cell-based assay. The sample with the highest signal (donor A) showed a complete neutralization capacity against both IFN-α2 and IFN-ω, while a second sample (donor B) showed neutralization against IFN-α2 only (Fig. 1b). Thus, 2/116 (1.5%) convalescent plasma samples from previously hospitalized donors had high titers of neutralizing auto-Abs against IFN-α2 and/or IFN-ω. Of the 4 unique donors, 2 had donated at multiple times and remained positive over all timepoints (Fig. 1c). It remains to be determined whether administration of convalescent plasma containing type I interferon autoantibodies has any detrimental effect to patients, particularly when diluted in the recipients’ blood volume. However, in our original report, plasma from most patients could neutralize the protective effect of IFN-α2 against SARS-CoV-2 in vitro even when diluted up to 10,000-fold [4]. In light of recent reports on varying efficacy of convalescent plasma for COVID-19 treatment, it is worth considering that in rare circumstances, the presence of autoantibodies in the donor plasma pool could explain some of the variance in clinical response. Ongoing variation in the viral spike protein may necessitate the continued use of convalescent plasma, due to the slower pace of generation of recombinant monoclonal antibodies and possibly the emergence of vaccineresistance [5]. While we previously found non-hospitalized donors to be anti-IFN-α2 negative [4], the data presented here suggest that COVID-19 convalescent plasma from previously hospitalized patients may require screening for type I interferon autoantibodies. Alternatively, COVID-19 plasma donation could exclude those donors with a record of severe to critical pneumonia until future studies can address whether transfer of autoantibodies constitutes harm to recipients. Mark S. Anderson and Joseph L. DeRisi contributed equally to this work.