Leveraging the United States Epicenter to Provide Insights on COVID-19 in Patients with Systemic Lupus Erythematosus.

Abstract

Objective:

To characterize patients with Systemic Lupus Erythematosus (SLE) affected by COVID-19 and to analyze associations of comorbidities and medications on infection outcomes.

Methods:

Patients with SLE and RT-PCR-confirmed COVID-19 were identified through an established New York University lupus cohort, query of two hospital systems, and referrals from rheumatologists. Data were prospectively collected via a web-based questionnaire and review of medical records. Baseline characteristics were obtained for all patients with COVID-19 to analyze risk factors for hospitalization. Data were also collected from asymptomatic patients and those with COVID-19-like symptoms who tested negative or were not tested. Statistical analyses were limited to confirmed COVID-19-positive patients.

Results:

A total of 226 SLE patients were included: 41 patients with confirmed COVID-19; 19 patients who tested negative for COVID-19; 42 patients with COVID-19-like symptoms who did not get tested; and 124 patients who remained asymptomatic without testing. Of those SLE patients with COVID-19, 24 (59%) required hospitalization, four required intensive care unit-level of care, and four died. Hospitalized patients tended to be older, non-white, Hispanic, have higher BMI, history of nephritis, and at least one comorbidity. An exploratory (due to limited sample size) logistic regression analysis identified race, presence of at least one comorbidity, and BMI as independent predictors of hospitalization.

Conclusion:

In general, the variables predictive of hospitalization in our SLE patients were similar to those identified in the general population. Further studies are needed to understand additional risk factors for poor COVID-19 outcomes in patients with SLE.

Introduction:

As New York City (NYC) navigates the ongoing coronavirus disease 2019 (COVID-19) pandemic, the ability to identify risk factors for morbidity and mortality, potential preventative measures, and effects of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) on the immune response and disease activity in those with autoimmunity has been limited by the initial absence of widespread testing. Several factors have been hypothesized to impact COVID-19 outcomes including comorbid diseases, race and ethnicity, immune abnormalities, and medication use. For example, conditions such as hypertension, chronic lung disease and chronic kidney disease have been associated with poor outcomes in COVID-19 (1, 2). Data from NYC revealed that Black/African American (AA) and Hispanic/Latino people were more likely to die of the virus (3). Abnormalities of the immune system including lymphopenia have been related to morbidity in COVID-19 patients, and a hyperinflammatory syndrome has been associated with acute worsening of disease manifestations (4, 5). Immunosuppressant medications have also been proposed as treatment in an attempt to forestall the potential cytokine storm associated with poor outcomes (6, 7) which raises the consideration that such therapies in patients with autoimmune diseases may potentially provide benefit rather than susceptibility (8).

Patients with systemic lupus erythematosus (SLE) represent a unique population in considering risk for COVID-19 with biologic, genetic, demographic, clinical and treatment issues at play. By the nature of their chronic inflammatory autoimmune condition and regular use of concomitant immunosuppressants, these individuals would traditionally be considered at high risk of contracting SARS-CoV-2 (9). However, it might be speculated that inherently elevated type I Interferon, characteristic of the majority of patients with SLE, confers a protective effect as a first line anti-viral defense (10). Additionally, hydroxychloroquine (HCQ) was considered potentially efficacious early on based on in vitro inhibition of viral replication as well as positive findings in very limited murine and human clinical studies (11–13); however, several recent studies have failed to demonstrate significant clinical effectiveness of HCQ in COVID-19 (14–17). Conversely, since SLE is most prevalent and severe in ancestrally African and Hispanic populations, this overlap with high-risk COVID-19 demographic groups also factors into risk assessment (18). Common SLE disease manifestations and comorbid conditions such as lung, kidney, and cardiovascular disease associate with poor outcomes in COVID-19 (19). Accordingly, this study was initiated to provide critical data needed to address the frequency and severity of COVID-19 in patients with SLE. To accomplish this goal, we launched longitudinal surveys and review of medical records of patients with SLE from the New York University (NYU) hospital systems and Bellevue Hospital Center with suspected or confirmed COVID-19, as well as SLE patients who have not developed symptomatic COVID-19.

Patients and Methods:

Study population

Patients with SLE and confirmed or suspected COVID-19 were recruited from several sources. The largest group were subjects from the established NYU Lupus Cohort (a prospective convenience registry of patients with SLE seen at NYU Langone Health and Bellevue Hospital Center since 2014) who had at least one outpatient visit and a blood sample collected in the last nine months (n = 176). The second group included referrals from NYU rheumatology providers as part of the Web-based Assessment of Autoimmune, Immune-Mediated, and Rheumatic Patients during the COVID-19 Pandemic (WARCOV) initiative (n = 33). The third group was obtained by ICD10 query of SLE (M32.0-M32.9) and testing for COVID-19 at Bellevue and NYU electronic medical record systems (n = 17); additional details are summarized in Figure 1. Participants were contacted from April 13, 2020 to June 1, 2020 via email or phone, with an Institutional Review Board (IRB)-approved questionnaire in their preferred language. This study was approved by the NYU and Bellevue Hospital IRB.

Figure 1. Flow diagram of included participants and sources of recruitment.

COVID, Coronavirus disease 2019; COVID-, patients who tested negative for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) by real-time reverse transcriptase polymerase chain reaction (RT-PCR); COVID+, patients who tested positive for SARS-CoV-2 by RT-PCR; WARCOV, Web-based Assessment of Autoimmune, Immune-Mediated, and Rheumatic Patients during the COVID-19 Pandemic.

Inclusion and exclusion criteria

All patients were age 18 or older and met at least one of the following criteria for SLE: 1) the American College of Rheumatology (ACR) revised classification criteria (20, 21); 2) the Systemic Lupus International Collaborating Clinics (SLICC) classification criteria (22); 3) the European League Against Rheumatism (EULAR)/ACR classification criteria (23); or 4) a previous rheumatologist’s diagnosis of SLE (if fulfillment of classification criteria for SLE was not feasible due to incomplete or unavailable data). Only English-, Spanish- or Mandarin-speaking patients were included in the study.

Study design, data collection, and outcome measures for hospitalized patients

A Research Electronic Data Capture (REDCap)-based questionnaire (24) was used to capture demographics, body mass index (BMI), comorbid illnesses that have been previously shown to increase the risk for poor COVID-19 outcomes (2, 25, 26), SLE medication use, patient’s self-reported SLE activity, potential COVID-19 related symptoms, the possibility of a confirmed COVID-19-positive contact, and whether the patient had been tested for COVID-19 by nasopharyngeal swab with the outcome. Patients were contacted weekly for 1–7 weeks to capture a change in COVID-19 and overall clinical status. Details of the hospital course were collected from hospitalized and deceased patients from their medical records. Additional data collected on hospitalized patients included highest level of care (i.e., need for admission to intensive care unit [ICU]), supplemental oxygen requirement, need for intubation and mechanical ventilation, extracorporeal membrane oxygenation (ECMO), and death. We also captured any documented venous thromboembolic events (VTE), need for renal replacement therapy, COVID-19-directed therapies, chest imaging findings, and laboratory testing including creatinine, absolute lymphocyte count, D-Dimer, ferritin, interleukin(IL)- 6, erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) levels. COVID-19 cases required a positive SARS-CoV-2 real-time reverse transcriptase polymerase chain reaction (RT-PCR) nasopharyngeal swab test (COVID-19+). Repeat testing was captured, particularly for inpatients with a high degree of suspicion of disease who initially tested negative. Given the difficulty in obtaining RT-PCR testing to detect COVID-19 in the outpatient setting during the early phases of the pandemic in NYC, we also report patients with COVID-19-like symptoms (as described in Table 1) who were not tested. Due to the overlapping symptoms of SLE and COVID-19, including fever (23), suspected COVID-19 patients were not combined with confirmed COVID-19 patients for the purpose of statistical analyses.

Table 1.

Characteristics of patients with SLE classified by COVID-19 status.

	SLE and Confirmed COVID-19 (RT-PCR: Positive) N = 41	SLE and COVID-19-like Symptoms (RT-PCR: Not Performed) N = 42	SLE and Tested for COVID-19 (RT-PCR: Negative) N = 19	SLE and No COVID-19-like Symptoms (RT-PCR: Not Performed) N = 124
Demographics
Age	47 ± 17.19	41 ± 12.68	45 ± 12.17	45 ± 13.67
Gender (N, %)
Female	38 (92.7%)	39 (92.9%)	19 (100%)	114 (91.9%)
Male	3 (7.3%)	3 (7.1%)	0	10 (8.1%)
Race (N, %)
White	11 (26.8%)	18 (42.9%)	8 (42.1%)	49 (39.5%)
Black	17 (41.5%)	12 (28.6%)	4 (21.1%)	29 (23.4%)
Asian	4 (9.8%)	3 (7.1%)	2 (10.5%)	27 (21.8%)
Other	9 (22.0%)	9 (21.4%)	5 (26.3%)	19 (15.3%)
Hispanic Ethnicity (N, %)	15 (36.6%)	18 (42.9%)	6 (31.6%)	29 (23.4%)
COVID-19 Symptoms*	N = 37
No Symptoms (N, %)	1 (2.7%)	0	9 (47.4%)	124 (100%)
Fever (N, %)	24 (64.9%)	19 (45.2%)	7 (36.8%)	0
Cough (N, %)	29 (78.4%)	22 (52.4%)	7 (36.8%)	0
Sore Throat (N, %)	12 (32.4%)	16 (38.1%)	6 (31.6%)	0
Diarrhea (N, %)	14 (37.8%)	10 (23.8%)	4 (21.1%)	0
Loss of Taste (N, %)	20 (54.1%)	9 (21.4%)	3 (15.8%)	0
Loss of Smell (N, %)	18 (48.6%)	10 (23.8%)	2 (10.5%)	0
Shortness of Breath (N, %)	24 (64.9%)	11 (26.2%)	6 (31.6%)	0
Known COVID Contact (N, %)	19 (51.4%)	11 (26.2%)	12 (63.2%)	21 (16.9%)
BMI† (kg/m2)	30.1 ± 9.69	27.1 ± 8.72	27 ± 5.7	25 ± 5.4
Perceived SLE Activity	N = 24
Complete Remission (N, %)	7 (29.2%)	6 (14.3%)	3 (15.8%)	39 (31.5%)
Mild (N, %)	11 (45.8%)	22 (52.4%)	9 (47.4%)	49 (39.5%)
Moderate (N, %)	5 (20.8%)	12 (28.6%)	5 (26.3%)	30 (24.2%)
Severe (N, %)	1 (4.2%)	2 (4.7%)	2 (10.5%)	6 (4.8%)
SLE Risk Factors
Prior LN (N, %)	14 (34.1%)	3 (7.1%)	5 (26.3%)	19 (15.3%)
APLS (N, %)	3 (7.3%)	1 (2.4%)	0	1 (0.8%)
Medications
Hydroxychloroquine (N, %)	32 (78.0%)	33 (78.6%)	15 (78.9%)	83 (66.9%)
Average Daily Dose (mg)	400	400	400	400
Systemic Steroids (N, %)	18 (43.9%)	5 (11.9%)	5 (26.3%)	16 (12.9%)
Average Daily Dose‡ (mg)	9 ± 8.71	16.7 ± 11.55 (N=3)	19.5 ± 23.61	11.7 ± 15.72
NSAIDs (N, %)	2 (4.9%)	6 (14.3%)	3 (15.8%)	4 (3.2%)
Immunosuppressants (N, %)	24 (58.5%)	24 (57.1%)	7 (36.8%)	56 (45.2%)
Mycophenolate Mofetil	10 (24.4%)	12 (28.6%)	2 (10.5%)	21 (16.9%)
Methotrexate	1 (2.4%)	2 (4.8%)	1 (5.3%)	9 (7.3%)
Azathioprine	4 (9.8%)	5 (11.9%)	0	9 (7.3%)
Belimumab	3 (7.3%)	4 (9.5%)	4 (21.1%)	13 (10.5%)
Cyclophosphamide	2 (4.9%)	0	0	2 (1.6%)
Rituximab	3 (7.3%)	1 (2.4%)	0	2 (1.6%)
Abatacept	1 (2.4%)	0	0	1 (0.8%)
acrolimus	5 (12.2%)	2 (4.8%)	0	2 (1.6%)
Tocilizumab§	1 (2.4%)	0	0	0
Other	3 (7.3%)	0	0	1 (0.8%)
Comorbidities
Pregnancy (N, %)	1 (2.4%)	0	1 (5.3%)	1 (0.8%)
Current Active Malignancy (N, %)	1 (2.4%)	0	0	1 (0.8%)
Organ Transplantation (N, %)	4 (9.8%)	1 (2.4%)	1 (5.3%)	2 (1.6%)
HTN, Not Controlled on Meds (N, %)	8 (19.5%)	3 (7.1%)	4 (21.1%)	10 (8.1%)
Diabetes Mellitus (N, %)	6 (14.6%)	0	1 (5.3%)	4 (3.2%)
COPD (N, %)	2 (4.9%)	0	0	2 (1.6%)
Congestive Heart Failure (N, %)	3 (7.3%)	0	1 (5.3%)	3 (2.4%)
Asthma (N, %)	6 (14.6%)	6 (14.3%)	5 (26.3%)	5 (4.0%)
Hospitalizations (N, %)	24 (58.5%)	0	2 (10.5%)	0

Values are expressed as N (%) for categorical variables and mean ± standard deviation (SD) for continuous variables.

^*Four patients who had tested positive for COVID-19 were hospitalized outside of NYU. Their COVID-related symptoms were not reported.

^†One patient who tested positive for COVID-19 was hospitalized outside of NYU and BMI was not reported. Additionally, one patient who tested negative for COVID-19 and two patients who had no COVID-19 symptoms and were never tested did not report BMI.

^‡Prednisone-equivalent dose.

^§Patient on tocilizumab for antibody-mediated rejection post-renal transplantation.

APLS, Antiphospholipid Syndrome; BMI, Body Mass Index; COPD, Chronic Obstructive Pulmonary Disease; COVID-19, Coronavirus Disease 2019; COVID-19-, negative testing for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) by real-time reverse transcriptase polymerase chain reaction (RT-PCR); COVID-19+, positive testing for SARS-CoV-2 by RT-PCR; HTN, Hypertension; LN, Lupus Nephritis; Meds, Medications; NSAID, Non-Steroidal Anti-Inflammatory Drugs; SLE, Systemic Lupus Erythematosus.

Statistical methods and analysis

Categorical variables were summarized by computing counts and proportions of patients (%). Continuous variables are expressed as mean ± standard deviation (SD), median with interquartile range (IQR), or range, as appropriate. Baseline characteristics were compared between the hospitalized and non-hospitalized (ambulatory) patients with COVID-19 in bivariate analyses using the Fisher’s exact tests for categorical variables and the two-sample T-test or Mann-Whitney U tests for continuous variables. An exploratory logistic regression analysis was also performed to identify potential independent predictors of hospital admission. Variable selection in the final model was based on both statistical significance (p<0.10, given limited sample size and power of the study; the impact of the variable on the estimated effects of the other covariates in the model) as well as clinical factors including whether there was prior evidence that the variable was an important predictor of COVID severity. All statistical analyses were performed using SAS version 9.4.

Results

Patient population

A total of 226 patients with SLE were included in this study. Of those, we identified 41 patients with confirmed COVID-19, 19 patients with negative RT-PCR for SARS-CoV-2 despite suggestive symptoms or known contact with a COVID-19 patient, 42 patients with COVID-19-like symptoms who did not get confirmatory testing, and 124 without any symptoms who were not tested for COVID-19 throughout the study period (Table 1). Most patients who completed the questionnaire perceived their SLE as being in remission or with mild activity, which was similar among the four groups. A diagnosis of SLE was confirmed by fulfillment of one or more classification criteria in 96% of patients, and only based on rheumatologists’ diagnoses in the remaining 4% of patients with SLE. Patients completed an average of 4 subsequent follow-up questionnaires (range 1–7) over the course of the study.

Clinical characteristics and outcomes of patients with SLE and confirmed COVID-19

Patients with SLE and confirmed COVID-19 (N=41) were predominantly female and encompassed the major racial/ethnic demographics characteristic of the NYC population at large (Table 1) (18). The most common symptoms of COVID-19 were cough (78.4%), fever (64.9%), and shortness of breath (64.9%). One patient had no specific symptoms suggestive of COVID-19 but felt generally unwell, prompting them to visit an urgent care center where they were diagnosed with COVID-19 pneumonia. One patient had an initial negative RT-PCR but underwent repeat testing given high index of suspicion, which returned positive. Four of the 41 COVID-19+ patients were subsequently tested for immunoglobulin G (IgG) antibodies to SARS-CoV-2 and all were positive.

At the time of development of COVID-19 symptoms, 34.1% of COVID-19+ SLE patients had a history of lupus nephritis (LN) and 7.3% had antiphospholipid syndrome (APLS). Most patients were on HCQ (78%), 43.9% were on systemic steroids and 58.5% were on at least one immunosuppressant, with mycophenolate mofetil (MMF, 24.4%) being the most common, followed by tacrolimus (12.2%). Comorbid conditions were common in the COVID-19+ cases, with 19.5% having uncontrolled hypertension, 14.6% with asthma, and a similar proportion of patients with diabetes mellitus. Of the COVID-19+ patients, 17 (41.5%) were treated as outpatient or in the emergency room, and the remaining 24 (58.5%) required hospitalization. Of those 24 patients, half required supplemental oxygen (52.7%) during hospitalization, 4 (16.7%) required admission to the ICU, and 3 (13%) required intubation and mechanical ventilation. One of the admitted patients presented with abdominal pain in the setting of COVID-19 and was found to have intestinal microthrombi in small bowel biopsy. One patient who was initially admitted and discharged for COVID-19 was subsequently readmitted for VTE complications and managed with anticoagulation therapy. One pregnant patient (third trimester) with respiratory symptoms was subsequently admitted for mild transaminitis and diagnosed with COVID-19 by RT-PCR. The patient did not have any adverse pregnancy outcomes. Four patients (16.7%) died of hypoxemic respiratory failure from COVID-19; of note, three of these patients had changed their code status to do not resuscitate/do not intubate (DNR/DNI), and one underwent intubation and renal replacement therapy during their ICU admission. The average age of the deceased patients was 72 ± 6.9 years and the majority (75% of patients) were non-white, female, and experiencing at least one comorbidity (uncontrolled HTN and LN were the most common, in 75% and 50% of patients, respectively). All four patients who died were on HCQ, two of them were on prednisone (≤7.5 mg/day), and one each were on cyclophosphamide and MMF. Additional characteristics of the admitted patients including laboratory and chest imaging results are provided in Table 2.

Table 2.

Characteristics of hospitalized SLE patients with RT-PCR-confirmed COVID-19.

	Hospitalized* N = 24
Details of Hospitalization (N, %)
Regular Floor	20 (83.3%)
ICU	4 (16.7%)
Oxygen/Ventilation Requirements (N, %)
Room Air (N = 19)	9 (47.4%)
Supplemental O2† (N = 19)	7 (36.8%)
Intubation/MV (N = 23)	3 (13.0%)
ECMO (N = 23)	0
Renal Replacement Therapy (N, %)‡
Required (N = 19)	2 (10.5%)
Not Required (N = 19)	17 (89.5%)
COVID-19-Directed Therapies (N, %)
Azithromycin (N = 20)	10 (50%)
Zinc (N = 20)	4 (20%)
HCQ added (N = 20)	3 (15%)
HCQ dose increased (N = 20)	2 (10%)
Steroid dose decreased (N = 20)	3 (15%)
Steroid dose increased (N = 20)	3 (15%)
Lopinavir/Ritonavir (N = 20)	1 (5%)
Tocilizumab (N = 20)	1 (5%)
Anakinra (N = 20)	1 (5%)
IVIG (N = 20)	1 (5%)
No treatment (N = 20)	2 (10%)
Laboratory Results§ (Median, Range)
ALC, 103μL (N = 15)	0.6 (0.04 – 2.3)
ESR, mm/hr (N = 7)	117 (44 – 168)
CRP, mg/L (N = 14)	65.4 (1.6 – 335.4)
Ferritin, ng/mL (N = 14)	740 (34.3 – 21845)
D-dimer, ng/mL (N = 13)	648.5 (113 – 27156)
IL-6, pg/mL (N = 5)	22.5 (<5 – 174.3)
Creatinine‡, mg/dL (N = 14)	1.7 (0.6 – 5.1)
Chest Imaging (N, %)
Abnormal Chest Imaging (N = 17)	14 (82.4%)
Normal Chest Imaging (N = 17)	3 (17.6%)
Death¶ (N, %)	4 (16.7%)

Values are expressed as % (N) for categorical variables and median (range) for continuous variables.

^*One patient was in an acute rehabilitation facility and another remained hospitalized as of July 2, 2020.

^†Supplemental oxygen administered via nasal canula, high-flow nasal cannula, simple or Venturi mask.

^‡One patient had end stage renal disease (ESRD) and was on hemodialysis at baseline (not included).

^§Laboratory results are peak values and nadir for ALC.

^¶Three patients were DNR/DNI.

ALC, Absolute Lymphocyte Count; CRP, C-Reactive Protein; COVID-19, Coronavirus Disease 2019; DNR/DNI, Do Not Intubate/Do Not Resuscitate; ECMO, Extracorporeal Membrane Oxygenation; ESR, Erythrocyte Sedimentation Rate; HCQ, Hydroxychloroquine; ICU, Intensive Care Unit; IL-6, Interleukin-6; IVIG, Intravenous Immunoglobulin; MV, Mechanical Ventilation; O_2, Oxygen; RT-PCR, real-time reverse transcriptase polymerase chain reaction.

A comparison between SLE patients treated in the ambulatory setting versus hospitalized cases is shown in Table 3. Hospitalized patients were slightly older than the non-admitted group (49 vs. 44 years old, respectively). The proportion of non-white and Hispanic patients were higher in the hospitalized compared to the ambulatory group (83.3% non-white and 41.7% Hispanics vs. 58.8% non-white and 29.4% Hispanics, respectively). Hospitalized patients were also more likely to have a history of LN (45.8% vs. 17.7%), APLS (12.5% vs. 0%), and one or more comorbid conditions such as asthma, chronic obstructive pulmonary disease, congestive heart failure, current active malignancy, diabetes mellitus, hypertension not controlled with current medications, organ transplantation, and pregnancy (62.5% vs. 29.4%). The variables that were initially considered for inclusion in the logistic regression model are listed in Table 3. In exploratory multivariable analyses, the independent predictor variables for hospitalization that were included in the final logistic regression model were race (Odds Ratio [OR] = 7.78 for non-white vs. white; 95% Confidence Interval [CI]: 1.13 to 53.58; p=0.037), the presence of one or more comorbidities (OR = 4.66; 95% CI: 1.02 to 21.20; p=0.047), and BMI (OR = 1.08 per increase in kg/m²; 95% CI: 0.99 to 1.18; p=0.096).

Table 3.

Comparison of hospitalized and ambulatory SLE patients with RT-PCR-confirmed COVID-19.

COVID-19+ Patients	Hospitalized (N = 24)	Ambulatory (N = 17)	P-value*
Age (years)	49.38 ± 17.81	43.65 ±16.26	0.30
Gender (N, %)			1.00
Female	22 (91.7%)	16 (94.1%)
Male	2 (8.3%)	1 (5.9%)
Race (N, %)			0.15
White	4 (16.7%)	7 (41.2%)
Non-White	20 (83.3%)	10 (58.8%)
Hispanic Ethnicity (N, %)	10 (41.7%)	5 (29.4%)	0.52
BMI (kg/m2)†	29.0 (25.7, 37.1)	27.4 (24.7, 30.7)	0.34
SLE Risk Factors (N, %)
APLS	3 (12.5%)	0 (0%)	0.25
History of LN	11 (45.8%)	3 (17.7%)	0.10
Medications (N, %)
Hydroxychloroquine	18 (75.0%)	14 (82.4%)	0.71
Systemic steroids	13 (54.2%)	5 (29.4%)	0.20
Prednisone-equivalent dose (mg/day)	3.9 ± 4.82	4.0 ± 9.93	0.96
Immunosuppressants‡	15 (62.5%)	9 (52.9%)	0.75
NSAIDs	1 (4.2%)	1 (5.9%)	1.00
Comorbidities ≥1 (N, %)	15 (62.5%)	5 (29.4%)	0.06

Values are expressed as % (N) for categorical variables and mean ± standard deviation (SD) or median (interquartile range [IQR]) for continuous variables.

^*Categorical variables compared using Fisher’s exact test; continuous variables compared using the two-sample T-test or Mann Whitney U Test. Age: T-test; BMI: Mann Whitney U Test (chosen by whichever gave more conservative p-value).

^†Median (IQR); N=23 for Hospitalized group.

^‡Immunosuppressants include non-biologic agents (azathioprine, cyclophosphamide, mycophenolate mofetil, mycophenolic acid, sirolimus, tacrolimus) and biologic agents (anakinra, abatacept, belimumab, rituximab, tocilizumab).

^§Comorbidities refer to at least one of the following: asthma, chronic obstructive pulmonary disease, congestive heart failure, current active malignancy, diabetes mellitus, hypertension not controlled with current medications, organ transplantation, and pregnancy.

APLS, Antiphospholipid Syndrome; BMI, Body Mass Index; COVID-19, Coronavirus Disease 2019; COVID-19+, positive testing for SARS-CoV-2 by polymerase chain reaction; LN, Lupus Nephritis; RT-PCR, real-time reverse transcriptase polymerase chain reaction SLE, Systemic Lupus Erythematosus.

Clinical characteristics and outcomes of patients with SLE and symptom-based suspected COVID-19

Table 1 reveals that an additional 42 patients had symptoms that could be consistent with COVID-19. The most common symptoms in these patients were cough (52.4%), fever (45.2%), and sore throat (38.1%). Eleven (26.2%) of these patients had known contact with a COVID-19 patient. These patients had similar gender and racial/ethnic demographics as the COVID-19+ SLE patients but were slightly younger (41 ± 12.7 years vs. 47 ± 17.2 years) and LN was less frequent (7.1% vs. 34.1%). While HCQ and immunosuppressant use was similar between the patients with suspected COVID-19 and the COVID-19+ cases (HCQ use: 78.6% vs. 78.0%, immunosuppressants use: 57.1% vs. 58.5%; respectively), a smaller percentage of patients in the suspected COVID-19 group were on systemic steroids compared to the COVID-19+ group (11.9% vs. 43.9%, respectively). Comorbid conditions were less common in the group of SLE patients with unconfirmed COVID-19. Two patients with COVID-19 symptoms who were initially unable to get confirmatory RT-PCR subsequently tested positive for IgG antibodies to SARS-CoV-2.

Clinical characteristics of SLE patients with negative SARS-COV-2 testing or asymptomatic for COVID-19

Ten SLE patients who had symptoms consistent with COVID-19 had a negative RT-PCR testing for COVID-19. The most common symptoms were cough (36.8%), fever (36.8%), shortness of breath (31.6%) and sore throat (31.6%). Additionally, 9 patients were asymptomatic but were tested due to a known COVID-19+ contact. Six patients had SARS-CoV-2 testing performed more than once given the high index of suspicion for COVID-19 (all negative). Two of those six patients were admitted: one with severe thrombocytopenia from thrombotic thrombocytopenic purpura, no respiratory symptoms, and an incidental finding of fever and lung infiltrates on chest imaging, who tested negative twice; and a second patient with end-stage renal disease, admitted with shortness of breath, who tested negative for COVID-19 three times. One patient who had negative RT-PCR testing for SARS-CoV-2 while asymptomatic but with a positive COVID-19 contact subsequently also tested negative for IgG antibody to SARS-CoV-2 at 39 days after RT-PCR testing.

Additionally, 124 SLE patients had no COVID-19 symptoms throughout the follow-up period and did not undergo RT-PCR testing. Of those, 16.9% had a known COVID-19-positive contact. One patient without any known contact with a COVID-19+ patient subsequently tested positive for IgG antibody to SARS-CoV-2. The demographics and medications of these patients are presented in Table 1.

Discussion

To our knowledge, our study represents the largest cohort of SLE patients with COVID-19 from the US epicenter of the pandemic. The impact of COVID-19 in the US, particularly in NYC, has been devastating, with over 205,000 cases and 17,000 confirmed deaths as of June 11, 2020 (27). In this observational prospective study, we describe 41 patients with SLE and RT-PCR-confirmed COVID-19. We also included 19 patients with negative SARS-CoV-2 testing, and 42 patients with COVID-19-like symptoms but without confirmatory testing. Our study is one of the first to date to follow a large cohort of SLE patients prospectively in an attempt to systematically identify emergent COVID-19 cases over time. With this approach, we additionally describe the demographic and clinical characteristics of 124 patients who remained asymptomatic and did not undergo RT-PCR testing for COVID-19 during the study period.

Data on patients with COVID-19 in the setting of autoimmune rheumatic diseases remain limited. However, multiple studies from China, Italy and the US have reported potential predictors of severe disease and mortality in non-rheumatic patients with COVID-19, including older age, male gender, BMI, diabetes, chronic kidney disease, hypertension, and heart failure (26, 28–30). In agreement with these observations, our study supports that the presence of at least one of these “classic” comorbidities and a higher BMI increase the risk of hospitalization in patients with SLE affected by COVID-19. Therefore, the variables predictive of hospitalization in our SLE patients were similar to those identified in the general population, as suggested by previous data (31). We have also identified non-white race as an independent predictor of hospitalization for COVID-19 in patients with SLE. Consistently, data from NYC suggests a disproportionate impact of COVID-19 in racial and ethnic minorities, with substantially higher rates of death among AA and Hispanic/Latino people compared to white people (3).

As SLE is relatively rare (32), studies that focus on this subset of patients consist mostly of small case series. In a French study of 17 SLE patients with COVID-19, all of whom were on long term HCQ, 82% were admitted to the hospital, 41% required admission to the ICU, and 14% of patients died from COVID-19-related complications (33). These findings parallel ours in that a relatively high percentage of SLE patients available for the study required hospitalization and died. Nevertheless, while 65% of patients in the French study required supplemental oxygen and half were intubated, these percentages were far lower for our patients, with only 13% requiring intubation and 36.8% necessitating supplemental oxygen; however, the former may reflect the fact that three of the patients who died in our study had DNR/DNI orders before death.

A letter from the COVID-19 Global Rheumatology Alliance (GRA) reported 80 COVID-19+ SLE patients, without any difference in the proportion of hospitalized patients based on antimalarial use (57% on antimalarials vs. 55% not on antimalarials) (34), which is consistent with our data. Additionally, 60% of the GRA-reported patients did not require supplemental oxygen, similar to our findings in hospitalized SLE patients with COVID-19. The largest publication to date from the GRA’s physician-reported registry described 85 patients with SLE, with a significantly higher percentage of SLE in hospitalized COVID-19 patients (17%) compared to those not requiring hospitalization (11%) (35). Based on a publication of ten COVID-19 cases and an additional eight presumed cases, the use of immunosuppressants on admission was similar between patients with mild versus severe disease (36). Similarly, our data did not show any association between the use of immunosuppressants and increased risk for hospitalization.

The emergence of venous thrombotic events has been described in up to 31% of critically ill patients with COVID-19 (37, 38). Interestingly, two of the 24 (8.3%) SLE patients with COVID-19 who required hospitalization were subsequently diagnosed with micro- and macro-thrombotic events, emphasizing the need for a high index of suspicion for this complication, especially in the setting of autoimmune rheumatic diseases, including SLE and APLS, where the risk of VTE is known to be higher than in the general population (39–41).

Early on during the COVID-19 pandemic, data suggested a potential beneficial role of HCQ and azithromycin in COVID-19 (13). The medications used to treat COVID-19 in our patients with SLE reflect the timing of COVID-19 diagnosis, when the use of HCQ and azithromycin was common practice in the inpatient setting (17). In contrast, recent COVID-19 clinical trials and cohort studies have started to show a favorable effect of anakinra, tocilizumab, and remdesivir in non-rheumatic patients with COVID-19 (42–47), which were less frequently used to treat patients in our study.

There are several limitations to our study, many of which relate to challenges of availability of testing and accuracy of detecting SARS-CoV-2. As NYC and the surrounding metropolitan area were the epicenter of the US COVID-19 pandemic, the initial ability to obtain timely testing was limited to patients who were sufficiently sick to require hospitalization (48). In our series, 42 patients had COVID-19 symptoms but were not tested. Data obtained from the COVID-19 outbreak on cruise ships revealed a high proportion of asymptomatic individuals with positive COVID-19 PCR results at the time of testing (49), emphasizing the potential for underestimating the incidence and prevalence of COVID-19. In addition, the false negative rate of testing approaches 30% (50), suggesting that some of our 19 patients who had one or more negative RT-PCR testing could have still had COVID-19. Fever is a common symptom of SLE, and for patients without SARS-CoV-2 testing, it is difficult to differentiate SLE activity from COVID-19. Therefore, we decided to strictly limit the definition of COVID-19 patients to those with a confirmatory SARS-CoV-2 RT-PCR test and exclude suspected COVID-19 cases from statistical analyses. It is acknowledged that despite NYC being the US epicenter of the pandemic, SLE approaches the definition of a rare disease (32) and thus the number of confirmed COVID-19 cases was limited. Referral bias is a concern given many of the SLE patients included in our study were referred from NYU rheumatologists for COVID-19-like symptoms, positive SARS-CoV-2 testing or were captured during their hospital admissions for COVID-19. However, the inclusion of patients from the NYU lupus cohort attempted to address this issue. Lastly, we could not account for frequency of SARS-CoV-2 relative to known exposure as many infected patients had no known COVID-19 contacts, and in NYC the estimated prevalence (based on preliminary state-wide antibody testing results) of COVID-19 is estimated to be 20% in the overall population and 27% in low-income minority communities (51). Because of these limitations, we decided to take a largely descriptive focus to our dataset and did not perform additional statistical analyses that could be confounded by the aforementioned factors.

Our study has several strengths. A systematic approach was used with the objective of contacting patients with SLE in a large, well characterized, multi-racial/ethnic cohort with questionnaires in three of the most common languages in NYC. We also describe cases that tested negative and provide data on patients who were not tested, with and without COVID-19-like symptoms.

In summary, our data suggest that patients with SLE and COVID-19 have a high rate of hospitalization but a similar mortality rate to the general population in NYC. Non-white race, BMI, and the presence of one or more comorbidities were identified as independent predictors of hospitalization in SLE patients who develop COVID-19, similar to the general population. There is insufficient evidence to conclude with our data that SLE-specific factors additionally contribute to the risk of hospitalization. Further studies are needed to understand additional risk factors for poor COVID-19 outcomes in patients with SLE.