We studied data from hospitalized adult patients with COVID-19 pneumonia treated subcutaneously with recombinant IFN-beta-1b (Betaferon) during admission at the Infanta Leonor University Hospital in Madrid, Spain, from the first wave of the pandemic (March to May 2020), in a retrospective manner. Based on the local treatment protocol at that time, when no drug had yet been proven efficacious, IFN-beta-1b was administered only to patients with severe respiratory failure and lack of initial response to the standard treatment (lopinavir/ritonavir, hydroxychloroquine, azithromycin).
We initially selected, from a cohort of 1,549 COVID-19 patients admitted to the hospital , 47 clinically severe patients who were treated with IFN-beta-1b (250 µg every 48 hours for up to 14 days) and who had an available plasma sample collected prospectively during hospitalization and stored in the hospital sample collection; this criterion was mandatory to guarantee the possibility to perform a test to determine the presence of anti-IFN autoantibodies.
All 47 patients were considered clinically severe at admission following the U.S National Institutes of Health (NIH) classification . The patients presented SpO2 <94% on room air at sea level, a ratio of arterial partial pressure of oxygen to fraction of inspired oxygen (PaO2/FiO2) <300 mm Hg, respiratory frequency >30 breaths/min, or lung infiltrates >50%. Patients who progressed to an ARDS were defined as critical following the NIH guidelines , with 200 mmHg <PaO2/FiO2 ≤ 300 mmHg to PaO2/FiO2 ≤ 100 mmHg, and required support with invasive or noninvasive mechanical ventilation at the ICU.
The primary outcome of this retrospective analysis of the IFN-beta-treated cohort was to evaluate the presence of anti-IFN autoantibodies. The secondary outcome was the effect of the presence of these antibodies on different clinical and laboratory variables and disease course.
Clinical data included demographics (age, sex, and ancestry), comorbidities (arterial hypertension, diabetes, dyslipidemia, overweight-obesity, and cardiac, renal, immunological, or tumor diseases), symptoms (fever and cough), analytics (D-dimer, C-reactive-protein, ferritin, and lymphocyte counts), treatment needs (corticosteroids, tocilizumab, and response to IFN-beta), oxygen needs (nasal cannula and mask or mechanical ventilation), and clinical outcomes (days of hospitalization, critical stage, and total deaths). The response to IFN-beta-1b was evaluated as a clinical improvement in no more than the median 9 days of hospitalization and no need for escalation in treatment after IFN-beta treatment .
The clinical data were collected retrospectively from the electronic medical records and entered into an anonymous electronic database (REDCap, Research Electronic Data Capture) . Different classes of variables were used for this study: a. response variables [presence of antibodies: dichotomous variable (presence/absence) and antibody titer: qualitative variable, factor]; b. explanatory variables: [survival: measured as time to death or discharge from hospital (censored data), requirements of O2, the severity of the disease, positive response to IFN (qualitative variables), and lack of response to IFN (assumed with no clinical improvement and the need for escalation in treatment with a bolus of corticosteroids or tocilizumab according to the local protocol or more than the mean 10 days of hospitalization after interferon administration]; and c. covariates: different demographic variables: comorbidities, laboratory values, symptoms, and treatments.
The biological blood samples were processed for the determination of anti-IFN autoantibodies (IFN-alpha, IFN-beta and IFN-omega) as described by P. Bastard et al, 2020 . In brief, 96-well ELISA plates (MaxiSorp; Thermo Fisher Scientific) were coated by incubation overnight at 4°C with 2 μg/ml rhIFN-α2 (Invitrogen), rhIFN-ω (Merck) or rh-IFN-β (Invitrogen). Plates were then washed (PBS/0.005% Tween), blocked by incubation with 5% nonfat milk powder in the same buffer, washed again, and incubated with 1:50 dilutions of plasma from the patients or controls for 2 h at room temperature (or with specific mAbs as positive controls). Each sample was tested once. Plates were thoroughly washed. HRP-conjugated Fc-specific (Fc, fragment crystallizable region) IgG fractions from polyclonal goat antiserum against human IgG (Nordic Immunological Laboratories) were added to a final concentration of 2 μg/ml. Plates were incubated for 1 h at room temperature and washed. Substrate was added, and the OD was measured.
Statistical analysis was carried out using IBM SPSS Statistics 26.0 (IBM Corp., Armonk, New York, USA). The variables that did not follow a normal distribution were expressed using the median and interquartile range (IQR). Nominal variables were expressed as numbers and percentages. Different models were used to study the association among the two response variables and the explanatory variables, as well as the covariates. To analyze the effect of dichotomous variables, the Chi2 test was used with a Markov chain correction for small sample sizes. The nonparametric Kruskal-Wallis test was used to explore the effect of continuous variables on the presence of antibodies. Likewise, the nonparametric Kaplan-Meier estimator was used to univariably analyze the survival function. Survival models with covariates were tested using the Cox proportional hazards model and the additive model of Aalen.
Informed consent was obtained orally when clinically possible. In the remaining cases, the informed consent waiver was authorized by the ethics committee. The study was approved by the Committee for Ethical Research of the Infanta Leonor University Hospital, code 008-20, and the Bellvitge University Hospital code PR127/20.