A Prospective Longitudinal Study Evaluating The Influence of Immunosuppressives and Other Factors On COVID-19 in Autoimmune Rheumatic Diseases
We conducted this study to identify the influence of prolonged use of hydroxychloroquine(HCQ), glucocorticoids (GC), and other immunosuppressants (IS) on the occurrence and outcome of COVID-19 in patients with autoimmune rheumatic diseases (AIRDs).
This was a prospective, multicenter, non-interventional longitudinal study across 15 specialist rheumatology centers. Consecutive AIRD patients on treatment with immunosuppressants were recruited and followed up longitudinally to assess parameters contributing to development of COVID-19 and its outcome.
COVID-19 occurred in 314(3.45%) of 9212 AIRD patients during a median follow up of 177 (IQR 129,219) days. Long-term HCQ use had no major impact on the occurrence or the outcome of COVID-19. Glucocorticoids in moderate dose (7.5-20mg/day) conferred a higher risk (RR = 1.72) of infection. Among the IS, Mycophenolate mofetil (MMF), Cyclophosphamide (CYC) and Rituximab (RTX) use was higher in patients with COVID 19. However, the conventional risk factors such as male sex (RR = 1.51), coexistent diabetes mellitus (RR = 1.64), pre-existing lung disease (RR = 2.01), and smoking (RR = 3.32) were the major contributing risk factors for COVID-19. Thirteen patients (4.14%) died, the strongest risk factor being pre-existing lung disease (RR = 6.36, p = 0.01). Incidence ((17.5 vs 5.3 per 1 lakh (Karnataka) and 25.3 vs 7.9 per 1 lakh(Kerala) ) and case fatality (4.1% vs 1.3 % (Karnataka) and 4.3% vs 0.4% (Kerala)) rate of COVID-19 was significantly higher (p < 0.001) compared to the general population of the corresponding geographic region.
Immunosuppressants have a differential impact on the risk of COVID-19 occurrence in AIRD patients. Older age, males, smokers, hypertensive, diabetic, and underlying lung disease contributed to higher risk. The incidence rate and the case fatality rate in AIRD patients is much higher than that in the general population.
A total of 9212 AIRD patients were recruited, the major disease subsets were RA(50.9%), SLE(15.4%), axial spondyloarthropathy (SpA)(9.1%) and psoriatic arthritis(PsA)(8%). The mean age of the cohort was 45.1 years (SD 14.3), 2% were in the pediatric age group (< 18yrs) and 77% were females. The median duration of underlying illness was 48 (22,96) months. Their baseline characteristics, IS administered and comorbidities are detailed in Table 1. During this study (median follow-up 177 days;(IQR)129,219), 314 patients (3.4%) were diagnosed with COVID-19 based on lab confirmation (RT-PCR and/ or RAT).
HCQ and COVID-19 Overall, during the study period, 57.4%, 68% and 88% of the total cohort, RA and SLE, were being treated with HCQ respectively. The mean dose of HCQ was 212 mg/day (SD 119). The median duration of HCQ use in the entire cohort was 12 months (IQR 3,39). In the overall cohort, HCQ use did not influence occurrence of COVID-19 (RR = 0.909, CI (0.715,1.154), p = 0.432) or mortality(p = 0.097) (Table 2, Fig. 1). In the subgroup analysis of RA and SLE, there was no independent impact of HCQ on the occurrence and outcome of COVID-19.
GC and COVID-19 (Table 1, Fig. 1): In the entire cohort, 3459 (37.7%) AIRD patients were using GC at the time of participation in this study. The use of GC was analyzed in 3 groups viz. <7.5 mg, 7.5-20mg, and > 20 mg, almost four-fifths (79.1%) of them were using a dose < 7.5 mg/day. In the univariate analysis, those on GC moderate dose category (7.5–20 mg), had the highest risk (RR = 1.72, C.I.(1.12,2.64), p = 0.01) of COVID-19, which was substantiated in multivariate analysis (RR 1.57, C.I.(1.003,2.47), p = 0.048). Current use of steroids trended towards moderate influence on mortality as well (RR 2.23, CI (0.87–5.71) p = 0.09)(Table 2).
Associations with other are, biologics
A higher association of COVID-19 positivity was noted with the use of cyclophosphamide (CYC) (p < 0.001) and mycophenolate (MMF) (p = 0.029) (Table 1). Amongst those on treatment with biologics, the risk was higher with rituximab (RTX) (p = 0.007). In the multivariate analysis, CYC [RR-4.2, CI (2.23, 7.91) p < 0.001] and RTX [RR 2.4, CI 1.35–4.32 p = 0.003] use were independent risk factors associated with increased risk of COVID-19. (Fig. 1, supp Table 1). However, MMF (P = 0.39), CYC (P = 0.36), and RTX (p = 0.58) use did not affect mortality (Table 2).
We found that a lesser proportion of RA patients developed COVID-19 compared to other AIRDs (RR-0.642, CI 0.51, 0.81, p = 0.041). Patients with systemic vasculitis had the highest risk of contracting COVID-19 (p < 0.001) (Table 1).
Comorbidities and other parameters
The parameters associated with COVID-19 positivity were older age (p = 0.025), male sex (p = 0.001), smoking (p = 0.002), comorbidities such as DM (p < 0.0001), HTN (p < 0.0001), presence of underlying lung disease such as interstitial lung diseases, asthma, chronic obstructive pulmonary disease (p < 0.0001) and smoking (p = 0.002). The relative risk of occurrence of COVID-19 with 95% CI is detailed in Fig. 1 (Supp Table 1). Multivariate analysis adjusted for age, gender, DM, HTN, lung involvement, and immunosuppression use, it was noted that male sex [RR-1.51, p = 0.001], the coexistent DM [RR-1.64, p = 0.001], underlying lung disease [RR-2.01, p < 0.0001], and smoking [RR-3.32, p < 0.0001] were the independent risk factors associated with the increased risk of COVID-19. The use of ACEi/ARBs had no significant impact on occurrence or outcome. Among the 314 COVID-19 patients, 13(4.14%) died. The factors associated with death are presented in Table 2. The pre-existing lung disease was the strongest risk factor associated with increased risk for mortality (RR 4.315, CI (1.416, 13.15)p < 0.01) apart from conventional risk factors such as smoking and coexisting DM.
Comparison of Incidence and mortality of COVID-19 with the general population
The incidence of COVID-19 in our cohort was significantly higher compared to the incidence in the general population (p < 0.0001) both in Karnataka and Kerala as depicted in Table 3. The case fatality rate was 4.5 times higher (4.1% in the AIRD cohort vs 0.9% in the general population) amongst the AIRD population.
Long term HCQ use had no significant impact on COVID-19 occurrence and mortality in AIRD patients while moderate doses of GC increased the risk of infection. AIRD patients who are elderly, male, smokers, hypertensive, diabetic and those with underlying lung disease have a higher risk of contracting COVID-19. The incidence rate is at least 3 fold higher and the case fatality rate is 4.6 times higher than that of the general population in the same geographic region during the same time period. Hence, this group of AIRD regardless of age and other comorbidities, merits first access to the various protective measures implemented against COVID-19.