SARS-CoV-2 antibody response after COVID-19 in patients with rheumatic disease

The impacts of rheumatic disease and immunosuppression on the development of antibodies to SARS-CoV-2 are unknown. A study of healthcare workers showed that detectable SARS-CoV-2 antibodies were associated with reduced risk of SARS-CoV-2 reinfection, and the robustness of this neutralising antibody response has implications for seroprevalence studies and vaccine efficacy.¹ While disease-modifying antirheumatic drugs (DMARDs) generally blunt the immune response to pathogens, immunosuppressive medications such as dexamethasone and baricitinib have efficacy in reducing the severity of COVID-19.^{2 3} Additionally, tumour necrosis factor inhibition has been proposed as a potential mechanism for enhancing germinal centre formation and antibody production in severe COVID-19.⁴ Understanding the SARS-CoV-2 antibody response after COVID-19 among rheumatic disease patients is therefore of particular interest.⁵

We examined the SARS-CoV-2 antibody response among patients with rheumatic diseases and past COVID-19 at the Mass General Brigham (MGB) health system in Boston, Massachusetts, USA. Patients with COVID-19 confirmed by positive PCR testing and rheumatic disease confirmed by electronic health record (EHR) review were identified as previously described.⁶ We extracted clinically obtained SARS-CoV-2 antibody results and other relevant variables from the EHR. This study was considered exempt by the MGB Institutional Review Board.

Out of 188 patients with PCR-confirmed COVID-19 and rheumatic disease, 13 patients had subsequent SARS-CoV-2 antibody testing (table 1). Of these, 2 had undetectable antibodies, 1 had variable results and 10 had positive antibodies. Of the two patients with negative antibodies, one patient had psoriatic arthritis treated with leflunomide and prednisone and had an uncomplicated COVID-19 course. The other patient had antineutrophil cytoplasmic antibody-associated vasculitis on rituximab, azathioprine and prednisone. This patient had negative SARS-CoV-2 antibodies between 28 and 216 days after COVID-19 and had a complicated course requiring intensive care unit admission. One patient with antiphospholipid syndrome on prednisone, cyclophosphamide, rituximab and eculizumab had initial positive antibodies 28 to 87 days after COVID-19. However, he had a negative antibody response by 107 days despite persistently positive PCR testing, phylogenetic analysis suggestive of persistent infection and viral evolution, and clinical concern for recurrent COVID-19, and he died from respiratory failure, as reported elsewhere.⁵

Table 1.

SARS-CoV-2 antibody test results in rheumatic disease patients with COVID-19 confirmed by PCR

Age, years	Sex	Rheumatic disease diagnosis	Rheumatic disease treatment	Timing of SARS-CoV-2 antibody test(s) relative to first positive COVID-19 PCR	SARS-CoV-2 antibody test result(s)	COVID-19 complications	COVID-19 pharmacologic treatment	COVID-19 clinical outcome
Negative/variable SARS-CoV-2 antibodies
48	Female	Psoriatic arthritis	Leflunomide 10 mg daily, prednisone 10 mg daily	T+177 days	Negative total antibody*	None	None	Fully recovered
62	Female	ANCA-associated vasculitis	Rituximab 1 g (started T–6 years, most recent dose T–149 days), azathioprine 100 mg daily, prednisone 7.5 mg daily	T+28 days† T+71 days T+111 days T+216 days	Negative IgG, negative IgM‡ Negative total antibody* Negative total antibody* Negative total antibody*	Hospitalisation with ICU admission Respiratory failure requiring oxygen therapy by high flow nasal cannula	Hydroxychloroquine, Remdesivir	Persistent cough (T+238 days). Oxygen requirement resolved by hospital discharge.
45	Male	Antiphospholipid syndrome	Prednisone 15 mg daily, cyclophosphamide 250 mg daily, rituximab 1 g (started T–5 years, most recent dose T–11 days), eculizumab 900 mg (started and most recent dose T-9 days)	T+28 days T+81 days T+87 days T+107 days	Positive IgM, negative IgG‡ Positive IgM, positive IgG‡ Positive IgM, positive IgG‡ Negative total antibody*	Hospitalisation with ICU admission Respiratory failure requiring mechanical ventilation; circulatory shock	Remdesivir, SARS-CoV-2 antibody cocktail (regeneron) (T+145 days)	Death (T+154 days)
Positive SARS-CoV-2 antibodies
26	Female	Systemic lupus erythematosus	None	T+1 hour T+7 days	Positive total antibody* Positive total antibody*	Hospitalisation with ICU admission TTP requiring plasma exchange and glucocorticoids	None	Recurrent TTP episode (T+58 days)
71	Female	Rheumatoid arthritis	None	T+58 days	Positive total antibody*	None	None	Fully recovered
73	Male	Psoriatic arthritis	Etanercept 50 mg weekly	T+60 days	Positive total antibody*	None	None	Fully recovered
54	Female	Systemic lupus erythematosus	Rituximab 720 mg (started T–86 days, most recent dose T–2 days)	T+60 days	IgG positive, IgM not performed‡	None	None	Fully recovered
63	Female	Systemic lupus erythematosus	Azathioprine 100 mg daily, belimumab 720 mg monthly (started T–336 days, most recent dose T–20 days)	T+88 days	Positive total antibody*	None	None	Fully recovered
55	Female	Sarcoidosis	None	T+93 days	Positive total antibody*	None	None	Fully recovered
52	Female	Rheumatoid arthritis	None	T+94 days T+210 days	Positive total antibody* Positive total antibody*	Hospitalisation without ICU admission Supplemental oxygen by nasal cannula	None	Fully recovered
68	Female	Polymyositis	Prednisone 6 mg daily, methotrexate 25 mg weekly	T+129 days	Positive total antibody*	None	None	Fully recovered
51	Female	Neurosarcoidosis	Methotrexate 15 mg weekly	T+155 days	Positive total antibody*	None	None	Fully recovered
72	Female	Psoriatic arthritis	Methotrexate 25 mg weekly	T+203 days	Positive total tntibody*	Hospitalisation without ICU admission; no oxygen requirement	None	Prolonged fatigue (T+262 days)

^*Measured with the Roche Elecsys assay, which reports the positivity of total SARS-CoV-2 antibody (IgM and IgG) and has 99.5% sensitivity at 14 days after COVID-19 infection.

^†T=time zero, defined as the date of the first positive COVID-19 PCR test.

^‡Measured with the Viracor Eurofins assay, which reports IgM and IgG antibody positivity to SARS-CoV-2. The sensitivity of the assay is unknown.

ANCA, antineutrophil cytoplasmic antibody; ICU, intensive care unit; PCR, polymerase chain reaction; T, time zero; TTP, thrombotic thrombocytopenic purpura.

The remaining 10 patients had detectable SARS-CoV-2 antibodies despite the presence of rheumatic diseases and/or the use of immunosuppressive medications, including prednisone, methotrexate, azathioprine, etanercept, rituximab and belimumab. The median time between SARS-CoV-2 PCR and antibody testing was 91 days (IQR: 60–146 days). Of these 10 patients, 8 patients had full recovery, 1 patient had persistent fatigue, and 1 patient with systemic lupus erythematosus (without prior haematologic involvement) had a complicated course with recurrent episodes of thrombotic thrombocytopenic purpura.

This case series of rheumatic disease patients with PCR-confirmed COVID-19 and clinically obtained SARS-CoV-2 antibody testing indicates that the majority of patients (10, 77%) developed detectable SARS-CoV-2 antibodies, which is reassuring. Three patients had negative or variable SARS-CoV-2 antibodies, and two of these patients had severe COVID-19. Three patients were on rituximab; two patients on rituximab for many years had undetectable circulating CD19+ B cells and undetectable or variable SARS-CoV-2 antibodies, while one patient who had recently started rituximab (flow cytometry not available) had detectable SARS-CoV-2 antibodies. As tests were obtained as part of routine clinical care at a tertiary care centre, generalisability may be limited, antibody titers and tests for neutralising antibodies are not available, and the timing of antibody testing relative to SARS-CoV-2 infection is variable. Further studies are needed to investigate the effects of specific rheumatic diseases and DMARDs on the efficacy and durability of the antibody response to SARS-CoV-2.