Environmental Triggers for Connective Tissue Disease

The Case of COVID-19 Associated With Dermatomyositis-Specific Autoantibodies

Maria De Santis; Natasa Isailovic, Francesca Motta; Caterina Ricordi; Angela Ceribelli; Ezio Lanza; Elena Azzolini; Salvatore Badalamenti; Antonio Voza; Carlo Selmi


Curr Opin Rheumatol. 2021;33(6):514-521. 

In This Article

Autoantibodies in COVID-19 Viral Infection: Our Experience

Viral agents may be involved in the onset of IIMs through several possible mechanisms, from changes in the host cellular proteins, no longer recognized by the host immune system, to stimulation of autoantibody production, carrying pathogenic idiotypes.[16] The infection and also its timing are relevant related to PM/DM disease onset, as most infections develop 3 months before the clinical onset of myositis[17,18] in particular in juvenile forms. Based on these elements, we investigated sera from 35 patients with a proven SARS-CoV-2 infection, i.e., a positive real time-polymerase chain reaction (RT-PCR) on a nasal swab or bronchoalveolar lavage, that were consecutively hospitalized in the Rheumatology Department (dedicated to low-intensity care of COVID-19 cases) at Humanitas Research Hospital between October 1st and November 30th, 2020. The work has been conducted in accordance of the Helsinki Declaration. Informed consent was obtained from all the patients. We collected data about demographic, clinical, and serologic characteristics, COVID-19 duration (i.e., date since the onset of symptoms), pharmacological treatments received during hospitalization, blood oxygenation at admission (i.e., the ratio between the partial pressure of oxygen and the fraction of inspired oxygen) and the highest level of oxygen support required, the need for intensive care unit admission, and mortality. Comorbidities and thromboembolic events (i.e., pulmonary embolism, deep and superficial vein thrombosis) were also recorded for all patients. In the clinical setting, local protocols allocated patients to receive dexamethasone if oxygen support was required (one patient received methylprednisolone, 40 mg) or to receive remdesivir if symptoms occurred less than 10 days before admission in the absence of renal or liver failure. No patient in this cohort was treated with other immune-modulators, including monoclonal antibodies.

Anti-nuclear antibodies (ANA) and cytoplasmic antibodies were tested by indirect immunofluorescence (IIF) on HEp-2 slides (INOVA Diagnostics, San Diego, CA, USA) as described,[19] and defined according to official standards (ICAP; https://www.anapatterns.org). Immunoprecipitation-Western Blot (IP-WB) was used to test for anti-MDA5, anti-MJ/NXP-2, antieIF2B, and anti-E2/E3 antibodies as previously described.[20,21] ELISA was used to test for anti-Scl70/TOPO1, anti-RNA Pol III, and anti-Jo1 (QUANTA Lite Jo-1, INOVA Diagnostics, San Diego, CA, USA) according to the manufacturer protocols.[22] Anti-RIG1 and anti-DHX58 were tested by ELISA using recombinant proteins (unpublished data), to evaluate the presence of these two different autoantibodies induced by the SARS-CoV-2 viral infection. RNA-IP was used to detect anti-Sjogren's syndrome-related antigen A (SSA)/Ro, anti-SSB/La, anti-RNPs (U2, U3, U1), anti-SRP, anti-Th/To, anti-Sm, anti-PL7, anti-PL12, and other antiaminoacyl tRNA synthetases as previously described.[23] Data were analyzed using the GraphPad Prism version 7.01 for Windows (GraphPad, San Diego, CA), and P values < 0.05 were considered statistically significant.

In our cohort of COVID-19 patients, 63% were males with a median age of 74, one patient had rheumatoid arthritis and one skin psoriasis. Patients came to the Emergency Department referring a median duration of COVID-19 symptoms of 4 days and the following hospital stay had a median duration of 15 days, with 28/35 patients (80%) requiring oxygen supplementation. Dexamethasone 6 mg intravenously for 10 days was administered to 27/28 patients who required oxygen support. Eleven patients also received remdesivir 200 mg IV for 5 days.

Results of ANA and cytoplasmic antibodies and specific autoantibodies are shown in Table 1, Table 2 and Table 3. We classified ANA and cytoplasmic positivity based on the predominant immunofluorescence staining, even though the coexistence of both patterns is possible. In fact, in 7/15 cytoplasmic positive patients (46.6%) we also observed a concomitant weaker nuclear fluorescence. Twenty/35 COVID-19 sera (57%) were ANA positive, in 95% at titers ≥1:160 most frequently with speckled (45%), nucleolar (30%) and homogeneous (15%) patterns. In 7/20 of ANA-positive patients (35%), specific autoantibodies were detected (Figure 1 panels a–g), in particular IP-WB confirmed 3/35 anti-MJ/NXP2 (Figure 1 panel h), 2/35 anti-RIG1, 1/35 anti-MDA5 (Figure 1 panel h), and one anti-Scl70/TOPO1 (Figure 1 panel g).

Figure 1.

Immunofluorescence staining pattern of specific autoantibodies from patients with COVID-19. Serum dilution, 1:160. (a) Anti-MJ/NXP2: few nuclear dots and cytoplasmatic reticular AMA-like. (b) Anti-MJ/NXP2: nucleolar clumpy and cytoplasmatic speckled. (c) Anti-MJ/NXP2: nucleolar punctate and cytoplasmatic speckled. (d) Anti-RIG1: weak nuclear homogenous and cytoplasmatic polar/Golgi-like patterns. (e) Anti-RIG1: weak nuclear speckled. (f) Anti-MDA5: nucleolar clumpy and cytoplasmatic reticular AMA-like. (g) Anti-Scl70/TOPO1: weak nuclear homogeneous. (h) IP-Western blot analysis of COVID-19 patients with autoreactivity for anti MJ/NXP2 (n = 3) and -MDA5 (n = 1) antibodies (CTR+: positive control, CTR–: negative control).

Eleven out of 15 COVID-19 sera (73%) with cytoplasmic pattern were also positive for antibodies against the E2/E3 subunit of the pyruvate dehydrogenase complex, known as part of the antimitochondrial antigenic complex. No reactivity for additional myositis-specific antibodies belonging to the antiaminoacyl tRNA synthetases family, myositis-associated antibodies, and antibodies linked to other rheumatological diseases was observed. The proportion of patients requiring oxygen support was significantly higher among ANA-positive patients (90% vs. 20% ANA-negative; P < 0.0001) and, as expected, remdesivir was also more frequently used in ANA-positive patients (35% vs. 26.6%; P = 0.025). ANA-positive patients had longer hospitalization and higher extent of lung involvement including both poorly and nonaerated patterns. Anti-E2/E3-positive patients were significantly older than negative patients, and despite the association of this autoantibody with primary biliary cholangitis, only 3/11 (27%) anti-E2/E3-positive patients had elevated transaminases. The anti-E2/E3 status did not influence lung scores at imaging, blood oxygenation, and oxygen requirement for SARS-CoV-2.