Clinical characteristics and outcomes of COVID-19 breakthrough infections among vaccinated patients with systemic autoimmune rheumatic diseases

SARS-CoV-2 vaccines reduce the risk of COVID-19 (1–3). However, some DMARDs, particularly glucocorticoids, methotrexate, mycophenolate mofetil, and rituximab, may blunt the immunologic response to COVID-19 vaccination (4). Little is known about the clinical efficacy of these vaccines at preventing COVID-19 infection in patients with systemic autoimmune rheumatic diseases (SARDs).

Mass General Brigham (MGB) is a large multicenter health care system in the Boston, Massachusetts, USA area. Patients with SARDs with a positive SARS-CoV-2 polymerase chain reaction (PCR) or antigen test between January 30^th, 2020 and July 30^th, 2021 at MGB were identified using diagnostic billing codes or referred by physicians, as previously described (5). From this cohort, we identified breakthrough infections in fully vaccinated patients, defined as a positive test ≥ 14 days after the final vaccine dose (6).

Of 786 SARDs patients with COVID-19, 340 occurred after the initial emergency use authorization for COVID-19 vaccination in the US. Of these, 16 (4.7%) were breakthrough infections (Supplementary Figure 1). Among the breakthrough infections, 12 (75%) were female, 11 (69%) were white, the median age was 50 years, and 12 (75%) had ≥ one comorbidity (Table 1). The most common SARDs included rheumatoid arthritis (6, 38%), inflammatory myositis (3, 19%), and systemic lupus erythematosus (3, 19%). Rituximab (5, 31%), glucocorticoids (5, 31%), mycophenolate mofetil or mycophenolic acid (4, 25%), and methotrexate (3, 19%) were the most frequent immunosuppressives recorded prior to first vaccine dose. One (6%) patient was on no DMARD or glucocorticoid at the time of their vaccine.

Table 1.

Patient characteristics, vaccination details, medication use, and infection details of COVID-19 breakthrough infections in fully vaccinated SARDs patients (N=16)

Patient Characteristics n, %
Female	12, 75
Age (median, IQR)	49.5 [38.0, 64.5]
Race
White	11, 68
Black	3, 20
Hispanic	2, 13
Rheumatic disease*
Rheumatoid arthritis	6, 38
Rheumatoid arthritis-associated interstitial lung disease	1, 6
Dermatomyositis	3, 19
Myositis-associated interstitial lung disease	1, 6
Systemic lupus erythematosus	3, 19
Ankylosing spondylitis	2, 13
IgG4-related disease	1, 6
Mixed connective tissue disease	1, 6
Hypocomplementemic urticarial vasculitis	1, 6
Psoriatic arthritis	1, 6
Comorbidities*
Hypertension	6, 38
Morbid obesity (BMI ≥ 40.0 kg/m2)	3, 19
Interstitial lung disease	2, 13
End stage renal disease	2, 13
Chronic Obstructive Pulmonary Disease	1, 6
Asthma	1, 6
Diabetes	1, 6
Obesity (BMI ≥ 30.0 kg/m2)	1, 6
Coronary artery disease	1, 6
Cancer	1, 6
Organ transplant	1, 6
Immunodeficiency	1, 6
Chronic neurological or neuromuscular disease	1, 6
Inflammatory bowel disease	1, 6
None	4, 25
Smoking status
Current	1, 6
Former	7, 44
Never	6, 38
Unknown	2, 13
Vaccination Details n, %
Vaccine Type
Pfizer-BioNtech	7, 44
Moderna	5, 31
Janssen/Johnson & Johnson	4, 25
Disease Activity at Vaccination
First vaccination
Active	5, 31
Inactive	11, 69
Second vaccination
Active	6, 38
Inactive	6, 38
Not applicable	4, 25
Rheumatic Disease Treatment Use Prior to First Vaccine Dose† n, %
Rituximab	5, 31
Glucocorticoids	5, 31
Mycophenolate mofetil or mycophenolic acid	4, 25
Methotrexate	3, 19
Tacrolimus	2, 13
Adalimumab	1, 6
Azathioprine	1, 6
Belimumab	1, 6
Hydroxychloroquine	1, 6
Intravenous Immunoglobulin	1, 6
Sulfasalazine	1, 6
Tocilizumab	1, 6
Ustekinumab	1, 6
None	1, 6
COVID-19 Infection Details n, %
Time (days) from second/final vaccine dose to infection (median, IQR)	54.0 [29.8, 79.0]
Infection acquisition
Close contact with confirmed or probable case of COVID-19	4, 25
Presence in a healthcare facility with COVID-19 cases	3, 19
Community acquired	3, 19
Unknown	8, 50
Treatment
No treatment/supportive care only	7, 44
Remdesivir	6, 38
Glucocorticoids	3, 19
Neutralizing monoclonal antibody	4, 25
Azithromycin	2, 13
Convalescent plasma	1, 6
Enrolled in clinical trial‡	1, 6
Any Symptoms
Yes	15, 93
No§	1, 6
Symptoms
Fever	9, 56
Cough	7, 44
Malaise	6, 38
Myalgia	5, 31
Rhinorrhea	5, 31
Headache	4, 25
Shortness of breath	4, 25
Sore throat	4, 25
Diarrhea/vomiting/nausea	3, 19
Anosmia	3, 19
Dysgeusia	3, 19
Chest pain	2, 13
Arthralgia	1, 6
Other	1, 6
Outcomes n, %††
Outpatient management alone	10, 63
Hospitalization	6, 38
Ventilation	1, 6
Death	2, 13
Unresolved symptoms	2, 13

^*Patients may have >1 SARD or comorbidity

^†One patient initiated methotrexate in between the first and second dose and one patient initiated rituximab between the second dose and infection

^‡NCT04501978; Phase 3 randomized, blinded, trial assessing treatments for hospitalized patients with COVID-19. Intervention arms: investigational drug + standard of care (Remdesivir) or placebo + standard of care (Remdesivir)

^§Diagnosed via pre-procedure PCR test, no reported symptoms in electronic health record

^††Symptoms were unresolved (one active infection recent diagnosed, one reporting ongoing symptoms: fatigue/malaise) in two (13%) cases.

Seven (44%) patients received the BNT162b2 (Pfizer-BioNtech) vaccine, five (31%) received the mRNA-1273 (Moderna) vaccine, and four (25%) received the AD26.COV2.S (Janssen/Johnson & Johnson) vaccine. The median time from final vaccine dose to infection was 54 days (Table 1). Among the 16 breakthrough infections, 15 (93%) were symptomatic and 6 (38%) patients were hospitalized during which four (25%) required supplemental oxygen and one (6%) required mechanical ventilation (Supplementary Table 1). DMARDs used prior to infection among hospitalized patients included rituximab (4, 25%) and mycophenolate mofetil or mycophenolic acid (2, 13%). Two (13%) patients died; both deceased patients had received rituximab and had interstitial lung disease.

In conclusion, a small portion of COVID-19 cases among SARDs patients in a large US healthcare system occurred among fully vaccinated patients. However, some patients required hospitalization that ultimately culminated in death. The most common SARDs treatments at the time of vaccination included those associated with blunted antibody responses to SARS-CoV-2 vaccination (4). These findings suggest that the blunted SARS-CoV-2 antibody response following COVID-19 vaccination in certain DMARD users may be associated with an increased risk of breakthrough infections that may be severe and even fatal. Of note, the blunted response observed among glucocorticoid users is dose-dependent, especially above 10mg/day of prednisone. Some DMARD users may require alternative risk mitigation strategies, including passive immunity or booster vaccines, and may need to continue shielding practices.

Our study has certain limitations. First, we did not study the risk of breakthrough infections among a cohort of vaccinated patients with a known denominator. Therefore, we cannot estimate the rate of breakthrough infections among SARDs patients. It is possible that the observed number of cases might be expected since no vaccine will prevent every infection. Second, the proportion of asymptomatic breakthrough infections observed in our study may be an underestimate because we only included patients who presented for testing. Third, we did not have SARS-CoV-2 antibody testing available for all patients and cannot rule out the possibility that SARD manifestations (e.g., interstitial lung disease) commonly treated with these medications contributed to the severity of the presentation.

In light of our findings, additional studies are urgently needed to estimate the risk of breakthrough infections among SARDs patients and to evaluate the efficacy of booster vaccines and other strategies for DMARD users with poor immunologic response to COVID-19 vaccination.