Participants originated from a large single center (Sao Paulo, Brazil) phase 4 controlled prospective study (no. NCT04754698, CoronavRheum) of immunogenicity and safety of two doses of Sinovac-CoronaVac vaccine in ARD patients and controls (CG)13. After applying the exclusion criteria, the final study groups consisted of 828 ARD patients and 207 controls vaccinated with 2 doses (Supplementary Figure 1). ARD group included patients with: 27.5% (n=228) rheumatoid arthritis, 24.9% (n=206) systemic lupus erythematosus, 23.7% (n=196) axial spondyloarthritis, 6.2% (n=51) primary vasculitis, 4.8% (n=40) idiopathic inflammatory myopathies, 4.3% (n=36) systemic sclerosis, and 4.3% (n=36) primary Sjögren’s syndrome. Regarding ARD current therapy at 6 weeks after the 2nd dose (D69) the most frequently used were: 62.0% (n=513) immunosuppressive drugs, 36.5% (n=302) prednisone and 35.7% (n=296) biologic therapy. ARD and CG groups were comparable regarding median current age (p=0.898), female sex (p>0.999) and Caucasian race (p=0.163) (Supplementary Table 1).
Anti-SARS-CoV-2 S1/S2 IgG assessed from D69 (6 weeks after vaccine second dose) to D210 are presented in Table 1. From D69 to D210, anti-S1/S2 IgG seropositivity rates reduced 23.8% in ARD [650 (78.5%) vs. 495 (59.8%), p<0.001] and 20% in CG [202 (97.6%) vs. 161 (77.8%), p<0.001], with moderate but lower IgG persistence in ARD compared to CG at D210 (p<0.001). IgG GMT from D69 to D210 declined significantly after the second dose in ARD [41.8 (38.0-46.0) vs. 26.1 (23.2-29.4) AU/mL, p<0.001] and in CG [99.6 (88.2-112.6) vs. 48.8 (40.3-59.0) AU/mL, p<0.001], with lower IgG levels in the former group at D210 (p<0.001) (Table 1). The decrease in IgG titer, calculated as 1 – Ln(IgG210/IgGD69), was significantly lower in ARD compared to CG [38% (95%CI 32-43%) vs. 51% (95%CI 43-58%), p=0.004].
From D69 to D210, NAb positivity declined 41% in ARD [539 (65.1%) vs. 318 (38.4%), p<0.001] and 39.7% in CG [181 (87.4%) vs. 109 (52.7%), p<0.001]. In the same period, a substantial reduction of 41% in NAb activity was observed in ARD [47.7% (IQR 15.0-75.1) vs. 15% (IQR 15.0-52.0), p<0.001] and of 39.7% in CG [66.1% (IQR 45.2-91.0) vs. 34.5% (IQR 15-76.8), p<0.001]. The decrease in NAb activity [1 - Ln(NAb activityD210/NAb activityD69)] in positive individuals at D69 was significantly higher in ARD compared to CG [54% (95%CI 51-57%) vs. 47% (95%CI 41-52%); p=0.024].
At D210, negative anti-S1/S2 IgG in ARD group was associated with older age (p=0.001), lower frequencies of systemic lupus erythematosus (p=0.005) and systemic sclerosis (p=0.024) and higher frequency of male sex (p=0.007) and rheumatoid arthritis (p<0.001). Regarding the influence of current therapy, patients seronegative for IgG at D210 were more often under prednisone (43.7% vs. 31.1%, p<0.001), anti-TNF (19.2% vs. 13.2%, p=0.012), abatacept (9.8% vs. 2.7%, p<0.001) and rituximab (3% vs. 1.1%, p=0.028), while hydroxychloroquine use was more frequent in ARD with positive anti-S1/S2 IgG (Table 2). Multivariate logistic regression analysis using IgG positivity at D210 as the dependent variable and as independent variables those with p<0.2 in univariate analysis, revealed that male sex (OR=0.56, 95%CI 0.40-0.79, p<0.001), prednisone use (OR=0.56; 95%CI 0.41-0.76, p<0.001), anti-TNF use (OR=0.66; 95%CI 0.45-0.96, p=0.031), abatacept use (OR=0.29; 95%CI 0.15-0.56, p<0.001) and rituximab use (OR=0.32; 95%CI 0.11-0.90, p=0.031) were significantly associated with absence of anti-S1/S2 IgG six months after vaccination in ARD patients.
For NAb analysis at D210, SLE diagnosis was less frequent in seronegative ARD patients (p=0.019) whereas biologic therapy (p=0.031), particularly abatacept use (p=0.018) was higher in patients without NAb (Table 2). After multivariate logistic regression analysis using NAb positivity as the dependent variable and as independent variables those with p<0.2 in univariate analysis, only abatacept use (OR=0.24; 95%CI 0.13-0.46, p=0.041) remained significantly associated with absence of NAb at D210 in ARD patients.
Analysis of incident cases of RT-PCR confirmed COVID-19 were evaluated from first dose to 10 days after the second dose (T1) and thereafter to 180 days after the second dose (T2). Evaluation of incident cases among ARD patients (n=1193) revealed n=79 cases (n=33 in T1 and n=46 in T2) of COVID-19 with n=14 hospitalizations (n=7 in T1 and n=7 in T2) and n=4 deaths (n=1 in T1 and n=3 in T2). The incident symptomatic COVID-19 cases reduced from 27.5 (95%CI 18.9-38.6)/100 person-years in T1 to 8.1 (95%CI 6.0-10.9)/100 person-years in T2, with an incidence rate decline of 19.4 (95%CI 12.7-26.1) cases/100 person-years (p<0.001). Likewise, in the same time periods hospitalizations due to COVID-19 decreased from 5.8 (95%CI 2.3-12.0)/100 person-years to 1.2 (95%CI 0.5-2.6)-/100 person-years, with an incidence rate decline of 4.6 (95%CI 1.8-7.4) hospitalizations/100 person-years (p=0.001).
Deaths due to COVID-19 occurred in n=4 ARD patients: n=1 in T1 and n=3 in T2 (p=0.694), after the 2nd dose of Sinovac-CoronaVac. The first death occurred in a 58 years-old female rheumatoid arthritis patient with bronchiolitis, diabetes mellitus, arterial hypertension, obesity, and hypothyroidism, under rituximab (last dose 5 months earlier) and prednisone 2.5 mg/day. The second death was on a 72 years-old male rheumatoid arthritis patient with chronic renal failure, pleural effusion, asthma, and dyslipidemia, under hydroxychloroquine and prednisone 2.5mg/day. Third death occurred in a 73 years-old female, with idiopathic inflammatory myopathy and interstitial lung disease, arterial hypertension, diabetes mellitus, and dyslipidemia, under mycophenolate mofetil, and another in a 65 years-old female patient, with granulomatosis with polyangiitis and arterial hypertension, under azathioprine and prednisone (2.5 mg/day).
Further comparative survival analysis of ARD and CG (p=0.152), with random selection of age and sex comparable subjects pointed that n=33 (ARD n=31 and CG n=2) RT-PCR confirmed COVID-19 incident symptomatic cases were reported during T1 (40 days) and n=52 cases (ARD n=44 in CG n=8) during T2 (180 days) evenly distributed along this period in ARD patients (Supplementary Figure 2).