Autoantibodies as potential prognostic factors for clinical outcomes related to COVID-19: a systematic review of inception prospective cohort studies with GRADE recommendations.

Abstract

This systematic review of inception prospective cohort studies aimed to investigate whether autoantibodies are potential prognostic factors for short- and long-term clinical outcomes of COVID-19. Searches were conducted in MEDLINE, EMBASE, AMED, GLOBAL HEALTH, and COCHRANE databases from 2019 to 2022. When possible, meta-analysis was conducted, otherwise findings from individual studies were reported using odds ratios (OR) with 95% confidence intervals (CI). Quality of evidence was summarized using the GRADE criteria. We identified 2292 references, 18 inception prospective cohort studies (3178 patients) were included in the systematic review, and 12 studies reached criteria for meta-analysis. Studies achieved, in general, low to moderate risk of bias. Moderate quality of evidence showed that anti-interferon (IFN) was associated with increased risk of severity (OR=7.75; CI=1.79-33.61) and mechanical ventilation (OR=4.19; CI=2.06-8.53), but not with COVID-19 mortality (OR=1.68; CI=0.63-4.44). Antiphospholipids were not associated with COVID-19 mortality (OR=1.42; CI=0.85-2.37; P=0.18; I2=3.21) nor with thrombosis risk (OR=1.41; CI: 0.71-2.8; P=0.33). Antinuclear antibody level was not associated with risk of mortality or severity (risk for mortality: OR=3.8; CI=0.78-18.6; P=0.1; I2: 32.3; severity: OR=1.74; CI=0.96-3.16; P=0.07). Evidence currently available is insufficient for a quantitative analysis of autoantibodies association with long COVID-19. Anti-IFN measurement should be considered in COVID-19 follow-up. In a population-based rational, optimized vaccination strategies should be considered for individuals with anti-IFN antibodies since it could represent a risk for a worse prognosis. High-quality prospective studies for short- and long-term disease effects and autoantibody evaluation are still needed.