4 Discussion
COVID-19, a pandemic respiratory infectious disease caused by severe acute respiratory syndrome coronavirus 2(SARS-CoV-2), could lead to autoimmune and autoinflammatory diseases10. Patients with ARD are associated with immune dysfunction, and often treated with immunosuppressive drugs, making them more susceptible to infection. However, our meta-analysis showed relatively little difference between the prevalence of SARS-CoV-2 infections in ARD patients(0.061) and the global prevalence of SARS-CoV-2 infections(0.077) and the prevalence of ARD patients even lower.
Among ARDs, the prevalence of PsA was the highest. The mucosal immune system is the largest constituent of the immune system, protecting against infectious threats at the primary internal surface11,12. SARS-CoV-2 infection occurs predominantly at the respiratory mucosal surfaces, and decreased mucosal immunity must be beneficial for virus infection13. The highest prevalence of PsA with red patches of skin topped with silvery scales suggests that PsA patients may have some defects in mucosal immunity, which is more prone to COVID-19 infections14.
Vasculitis patients infected with COVID-19 had the highest mortality and hospitalization rates. COVID-19 made a life-threatening escalation from Th2 immune response to type 3 hypersensitivity with the subsequent deposition of antigen-antibody complexes, particularly inside the walls of blood vessels, to such an extent as to generate a systemic vasculitis in the context of complex immune disease15,16. According to Gell and Coombs classification, type 3 hypersensitivity occurs when excess or slight excess of soluble antigen leads to the accumulation of immune complexes, and the innate immune system cannot fully clear from the circulation. These antigen-antibody complexes precipitate in tissues, especially in blood vessels, and induce a severe inflammatory state through the action of complement anaphylaxis toxins(C3a and C5a), which in turn stimulates the release of histamine by mast cells and the recruitment of phagocytes17. Histopathologically, the result of this process in the vascular wall is acute necrotizing vasculitis, accompanied by neutrophil infiltration, karyorrhexis, and fibrinoid necrosis18. Platelet aggregation is associated with hypercoagulability and thrombosis described during COVID-1919. To prevent C3 activation, anaphylatoxin in this process through specific inhibitors, like compstatin-based AMY-101, provides effective therapeutic results15,20.
Among the ARD patients with COVID-19 infection, SLE patients had the highest ICU admission rate. They had higher rates of hypertension, diabetes mellitus, obesity, and sedentary lifestyle, which were significant risk factors for hospitalization and mortality in COVID-1921. Besides, data from the C19-GRA registry suggested that long-term use of GC was associated with poorer outcomes5,22. GC is frequently used to treat SLE, potentially making these patients vulnerable to a more severe COVID-19 disease course.
The influence of drugs on the COVID-19 infection rate of patients with ARDs is one of the biggest concerns of medical care. We were surprised that the prevalence of SARS-CoV-2 infection in patients with ARDs using GC or csDMARDs or bDMARDs was lower than the global prevalence of COVID-19 infection, indicating that these treatments did not necessarily increase this infection rate. Among all ARD patients infected with COVID-19, patients mainly treated with GC had the highest prevalence. GCs have strong anti-inflammatory effects and induce the apoptosis of lymphocytes. These immune inhibitory effects prevent lethality by excessive inflammation, but simultaneously increase the susceptibility to infection23,24.
Compared with the clinical outcomes of all COVID-19 patients with ARDs treated with several therapies, we found patients using bDMARDs or anti-TNF therapy had lower hospitalization rates, ICU admission rates, and mortality. Besides, bDMARD monotherapy was associated with a lower odds of hospitalization regression coefficient, and largely driven by anti-TNF therapies. Mechanistic studies have suggested that upregulated TNF-α may lead to inflammatory cell death, aberrant germinal center formation, and less robust humoral immune responses in fatal COVID-19, providing a potential biologic rationale for the protective effect of TNF inhibitors5,25-28. In vitro studies have demonstrated that IL-6 and TNF-α have been regulated by the recombinant protein S of SARS-CoV 2002, indicating that TNF-α inhibitors or the IL-6 inhibitors could decrease the cytokine storm in COVID-19 patients29,30. Anti-TNF therapy has been proposed as a potential treatment for the hyperinflammatory phase of severe COVID-1931-33.
We found that GC monotherapy was associated with higher odds of the hospitalization regression coefficient. Patients using csDMARDs had higher hospitalization rates, ICU admission rates, and mortality, but patients using GC had a higher hospitalization rate, higher ICU admission rate, and lower mortality. In the included studies, the csDMARDs drug used by most patients was hydroxychloroquine(HCQ). The World Health Organization SOLIDARITY trial reported no clinical benefit from HCQ for hospitalized patients with COVID-1934. Abundant evidence from numerous RCTs revealed HCQ does not prevent SARS-CoV-2 infection or improve outcomes in mild, moderate, or severe COVID-1935. HCQ use may even be associated with lower survival36. Therefore, csDMARDs, especially HCQ, are not suggested in the treatment of COVID-1925. Corticosteroids downregulate the release of cytokines, reducing the damage caused by cytokine storms25. GC is effective in patients with severe COVID-19-related respiratory failure, but may not have a benefit in patients with early or mild disease. The Infectious Diseases Society of America(IDSA) guidelines recommend dexamethasone in patients with COVID-19, hypoxemia, and/or critical illness37.
The effects of drugs used on clinical outcomes varied across ARDs. SARS-CoV-2-infected SLE patients with GC treatment had the highest hospitalization rates. sGC use(prednisone > 5 mg/day) in SLE is a risk factor for severe COVID-19 requiring hospitalization38. SARS-CoV-2-infected RA patients with csDMARDs or GC treatment had higher hospitalization rates, and those using GC had the highest mortality. Continuous GC users, on the one hand, inhibit the immune response and delay the clearance of the pathogen. They suppress the host inflammatory response, which in the case of viral infections of the respiratory tract is the primary responsible for lung damage and ARDs39. Clinic evidence point to a predominantly negative effect of continuous corticosteroids in managing this type of infection39.
Our study had several limitations. Meta-analyses of studies regarding the prevalence and clinical outcomes of COVID-19 had considerable heterogeneities. The cause of this heterogeneity could be explained by the differences in study size, including different diseases and study locations. Thus, we performed a subgroup analysis to assess the prevalence and clinical outcome of COVID-19 for each disease and performed a regression. Second, we could not obtain the specific drug use of each patient, and most patients used multi-drug combination therapy, which led to biases in our analysis of the impact of drugs on the prevalence of COVID-19 and clinical outcomes. Third, the sensitivity of RT-PCR for SARS-CoV-2 from the nasopharyngeal swab is roughly 70%40,41. Fourth, only published studies were included in the meta-analysis, and unpublished articles were excluded; publication and study selection bias may have affected our results. Therefore, further studies are needed to verify our results.