Management and outcome of COVID-19 in CTLA-4 insufficiency

With more than 6 million deaths and 700 000 000 cases worldwide, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has affected patients across all disease states, including those with rare inborn errors of immunity (IEIs), a heterogenous group of >470 genetic disorders characterized by susceptibility to infection and/or immune dysregulation. Clinicians caring for patients with IEIs face the challenge of making recommendations based on limited data. One multicenter study reported the clinical outcomes of 94 patients with IEI and COVID-19 but included a highly heterogenous group of genetic etiologies, limiting conclusions that can be made for specific genetic defects.¹ One of the largest cohorts examining the effects of COVID-19 on a single gene disorder included 22 patients with the immune dysregulatory disorder autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) who harbor neutralizing autoantibodies against type 1 interferons (IFNs), which are known to be important for early viral immunity^2-4; patients with APECED were found to have high rates of severe disease and mechanical ventilation. In patients with APECED and in the general population, autoantibodies against type 1 IFNs have been associated with severe COVID-19.² However, for most monogenic IEI, including primary immunoregulatory disorders driven by T-cell and B-cell autoimmunity, the risk of severe COVID-19 or the presence of autoantibodies against type 1 IFNs have not been established.

CTLA-4 insufficiency is an autosomal dominant, monogenic IEI characterized by multiorgan autoimmunity and immune dysregulation with variable penetrance.^5-7 CTLA-4 is an inhibitory immune checkpoint receptor constitutively expressed by regulatory T cells (T_regs), competing with CD28 for its ligands CD80/CD86 on antigen-presenting cells, which play a pivotal role in establishing immune tolerance.^8,9 Genetic defects in CTLA-4 lead to decreased T_reg numbers and suppressor function, T-cell infiltration of target organs, hypogammaglobulinemia, lymphopenia, natural killer cell dysfunction, autoantibodies against several tissues, and increased susceptibility to bacterial, viral, and parasitic infections.^5,6,10,11 The risk of severe COVID-19 in CTLA-4 is currently unknown. Furthermore, both patients and providers are often concerned about the possibility of disease exacerbations triggered by SARS CoV-2 vaccination, which has not yet been studied in this patient population. Given the susceptibility to autoimmunity and lymphopenia in CTLA-4 insufficiency, we hypothesized that our patients would develop higher rates of severe disease than the general population and that disease severity could be correlated with the numbers of autoantibodies against type 1 IFNs. Here we report COVID-19 severity, clinical presentation, management, reinfection rates, incidence of severe reactions to SARS-CoV-2 vaccination, and antibody responses of patients with CTLA-4 insufficiency. In addition, we evaluated the presence of autoantibodies against type I IFNs at baseline.

Patients were followed up in accordance with National Institute of Allergy and Infectious Disease Institutional Review Board–approved protocols. All patients were diagnosed with COVID-19 by their local providers using polymerase chain reaction or antigen testing. Data were obtained via patient interviews and chart reviews and prospectively collected under our natural history protocol. Additional data were retrospectively collected and analyzed for the purpose of this study. The severity of COVID-19 was assessed as asymptomatic, mild, moderate, severe, or critical illness based on published guidelines.¹² Screening for anti-IFN autoantibodies in the plasma samples was performed using a particle-based assay. Briefly, we covalently coupled differentially fluorescent magnetic beads to recombinant human type 1 IFNs and incubated them for 30 minutes with 1:100 diluted serum or plasma samples. Subsequently, the samples were incubated with PE-labeled goat anti-human immunoglobulin G antibody for 30 minutes, analyzed using a BioPlex X200 instrument, and the mean fluorescence intensity was quantified. The methods have been previously described in detail by Shaw et al.¹³ Patients with CTLA-4 insufficiency who did not contract COVID-19 during the study period, healthy controls, and patients with APECED were used as controls. All plasma samples predated the SARS-CoV-2 infection and were retrieved from the National Institute of Allergy and Infectious Disease/National Institutes of Health biobank. One-way analysis of variance was conducted to compare 3 or more groups, followed by post hoc Tukey tests. Student t or Mann Whitney tests was used to compare the groups. Fischer exact test was used to compare categorical variables.

Twenty-two patients (17 residing in the United States [P1-P17], 4 in Germany [P18-P21], and 1 in Uruguay [P22]) developed COVID-19 from June 2020 to July 2022. The patients were aged from 8 months to 54 years (mean age, 31 years), 11 of 22 were female, and all had genotypically confirmed CTLA-4 insufficiency (Table 1). Twenty patients had 2 or more disease manifestations of CTLA-4 insufficiency. P14 and P17 were diagnosed based on an affected sibling but were asymptomatic. Baseline treatment for CLTA-4 insufficiency included mammalian target of rapamycin inhibitors (8/22), immunoglobulin replacement (12/22), abatacept (8/22), and/or steroids (4/22). Most patients were lymphopenic at baseline before contracting COVID-19 (median absolute lymphocyte count, 1080 cells/mL). Although variant strain identification was available for only 1 patient (P4, who tested positive for Omicron during the second COVID-19 infection), we plotted each COVID-19 episode in relation to the Center for Disease Control variance surveillance data for all patients who resided in the US at the time of infection (Figure 1). Approximately half of the patients (10/22) had received at least 1 dose of SARS CoV-2 vaccine at the time of their first COVID-19 episode (Figure 1). The most common symptoms were fever (11/22), nasal congestion (11/22), and cough (10/22) (Table 1). Most patients had mild disease (20/22) and were managed as outpatients. Two patients had moderate or severe disease: P3 was hospitalized for COVID-19 pneumonia requiring supplemental oxygen, and P19 reported dyspnea without requiring oxygen supplementation. No patient required critical care unit admission or mechanical ventilation. Baseline sera were available for 11 patients (including P3), and none had neutralizing autoantibodies against IFN-α, IFN-β, or IFN-ω before contracting SARS-CoV-2 (Figure 2).

Baseline treatment for CTLA-4 insufficiency was continued in all patients, except for P10, in whom ustekinumab was held during COVID-19 infection. Eleven patients received 1 of the 3 monoclonal antibodies against the SARS-CoV-2 spike protein (anti-S): casivirimab + endevimab (7), bamlavinimab (2), or sotrovimab (2). Four patients received remdesivir, and 1 was treated with dexamethasone (P3) (Table 1). Seven patients received no treatment other than supportive care. The number of baseline clinical manifestations related to CLTA-4 insufficiency was lower in patients who received supportive care only (2.5 vs 6; P = .005). Initiation of SARS-CoV-2–directed treatment after symptom onset was 5 days for any agent and 4.2 days for anti-S monoclonal antibodies. Most patients with interstitial lung disease (7/8) received SARS-CoV-2–directed treatment and had a mild course, despite the presence of chronic lung disease. All patients showed complete resolution of symptoms (median duration, 7.5 days; range, 2-30). P13 had prolonged fatigue, which resolved several months after the initial infection. There were no cases of multisystem inflammatory syndrome. Three patients in whom serial SARS-CoV-2 polymerase chain reaction testing was performed (P1, P2, and P3) showed prolonged viral shedding beyond 30 days.

Nine patients developed a second COVID-19 episode (P1, P4, P7, P10, P12, P15, P18, P20, and P21) and 2 patients (P10 and P18) developed a third (Figure 1). The median symptom duration for repeat infection was 8.5 days (range, 3-30), and all cases were mild. P10 developed Long COVID-19, with persistent fatigue, headaches, brain fog, memory loss, leg and joint pain, dyspnea with exertion, and worsening of colitis after his third COVID-19 episode. P18 had an episode of immune thrombocytopenic purpura after the second COVID-19 episode, which responded to rituximab. Of the 9 patients who had repeated infections, 4 were not vaccinated (P1, P4, P7, and P20).

Seventeen patients were vaccinated (BNT162 = 10, messenger RNA 1273 [mRNA-1273] = 4, ChAdOx1 in combination with BNT162 = 1, CoronaVac in combination with BNT162 = 1, unknown = 1), and 6 of these patients (P8, P11, P16, P19, P21, and P22) received 3 or more doses. For all vaccinated patients, the median number of doses was 2. There were no cases of vaccine-associated myocarditis, severe vaccine-associated adverse effects, or disease flares attributable to the mRNA vaccination. P21 had an autoimmune hemolytic anemia flare within 2 weeks of receiving the first dose of ChAdOx1. Serologies were available for 16 patients. The median time between anti-S serology and the most recent SARS-CoV-2 vaccination or COVID-19 episode was 130 days (interquartile range, 85-251), and the median anti-S titers were 2090 IU/dL (interquartile range, 46-7583 IU/dL) (Figure 3). In patients receiving immunoglobulin replacement therapy (IGRT) (n = 10), the median anti-S titer was 349 IU/dL, compared with 2594 IU/dL in patients not receiving IGRT (n = 6; P = 0.15) (Figure 3). In both groups, the median time between serologies and the most recent vaccination or COVID-19 was similar (118 days vs 132 days; P = 0.7). P16 had high anti-S titers after receiving tixagevimab/cilgavimab and was not included in the analysis. Pre- and postvaccination serologies were only available for 2 patients receiving IGRT. Both patients had increased anti-S titers after vaccination (6.8-6348 IU/dL 112 days after vaccination in P2 and 755-6632 IU/dL 118 days after vaccination in P8).

Despite the high propensity for exuberant inflammatory responses in patients with CTLA-4 insufficiency, resulting from impaired T_reg function and immune dysregulation, our findings indicate that most patients with CTLA-4 develop a mild course of COVID-19. Although a significant number of patients had potential risk factors for severe disease, such as having multiple CTLA-4-associated comorbid conditions, chronic lung disease (interstitial lung disease), unvaccinated status, and contracting SARS-CoV-2 when Alpha and Delta variants were predominant, only 2 patients required hospitalization, and both did not necessitate mechanical ventilation; 90% of the patients were treated as outpatients, with a time to symptom resolution similar to that of the general population. Although the outcomes in most patients with CTLA-4 insufficiency seem reassuring, it is important to highlight that most patients were young adults, had immediate access to care, were vaccinated at the time of infection, underwent testing quite early in their disease course, were closely followed up by their local providers and our CTLA-4 clinical team, and continued their baseline immunomodulatory treatment during their COVID-19 course. Furthermore, patients who received specific treatment for SARS-CoV-2 had twice the baseline comorbidities of those who received supportive care only, suggesting that the decision to start such treatment might have been influenced by the disease burden at baseline. Anti-S protein antibodies, when started early, have resulted in a reduction in hospitalization risk, especially in patients at high risk.^15-18 Although there are no data on the use of these monoclonal antibodies or convalescent plasma in CTLA-4 insufficiency, we believe that our patients were at high risk and could benefit from treatment, and it is possible that treatment ameliorated the disease course in some patients. It is also important to note that although the disease course in most patients was mild, 3 patients had complications from their disease, including hypoxemia, immune thrombocytopenic purpura (responsive to rituximab), and prolonged COVID-19 with difficult-to-treat colitis.

Our observations also suggest that CTLA-4 insufficient patients can safely receive mRNA vaccines, with no severe adverse effects reported in our cohort. Although unvaccinated patients were more likely to develop COVID-19 and 2 unvaccinated patients developed COVID-19-associated complications (severe disease in P3 and long COVID-19 in P10), no conclusions regarding vaccine protection can be made because of possible confounders (eg, dominant SARS-CoV-2 variant at the time of infection, use of masks and other behavioral factors, etc). Although our study was not powered or designed to detect differences in anti-S titers between patients on immunoglobulin replacement therapy and those without, the median anti-S titers were 83% lower in patients receiving IGRT, with a similar time from COVID-19 vaccination or infection in both groups. Finally, in contrast to patients with autoantibodies against IFN-α, IFN-β, and IFN-ω and life-threatening COVID-19 infection,^2,4 our cohort of CTLA-4 insufficient patients did not have neutralizing autoantibodies against type 1 IFNs. Whether there is a potential role for autoantibodies against type 1 IFNs for risk stratification in patients with IEI and propensity for autoimmunity warrants further studies.

Our report represents an observational retrospective study of a rare disorder of genetically determined immune dysregulation and autoimmunity. Our relatively small sample size, differences in underlying disease severity, and immunomodulatory treatment at the time of SARS-CoV-2 infection and vaccination represent challenges in establishing firm conclusions regarding the management of COVID-19.

With this study, we seek to add to the scarce literature on the outcomes and management of patients with underlying IEIs and SARS-CoV-2 infection and help providers caring for patients with CLTA-4 insufficiency to better understand COVID-19 disease outcomes. The management and outcomes of our CTLA-4 insufficient cohort could potentially be extrapolated to patients with cancer who are at risk of developing autoimmune complications while being treated with the CTLA4-blocking checkpoint inhibitor ipilimumab.¹⁹ 

The authors thank the families and patients who participated in this study. This study was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health Intramural funds, and National Institutes of Health Clinical Center Bench-to-Bedside Program grant awarded to G.U. for the project titled “Deciphering Immune Dysregulation due to Germ line CTLA4 mutations” (480707). This work was also supported by intramural research funds from German Research Society grant SFB1160/2_B5 and German Ministry of Education and Research grant GAIN01GM1910A.

Contribution: S.O., M.S.A., and G.U. designed the research, performed the research, contributed vital new reagents or analytical tools, analyzed data, and wrote the manuscript; L.R. contributed vital new reagents or analytical tools; and A.R., K.H., J.A.L., B.L.W., D.S., M.K., B.G., and M.S.L. performed the research, analyzed data, and wrote the manuscript.

Conflict-of-interest disclosure: J.A.L. has received institutional funding from DBV Technologies, Aimmune, Regeneron, and Leadiant Biosciences. B.G. has received research funding and clinical trial funding from Bristol Myers Squibb, the manufacturer of Orencia (abatacept), however, not affecting any of the data presented in this manuscript. The remaining authors declare no competing financial interests.

Correspondence: Gulbu Uzel, Laboratory of Clinical Immunology and Microbiology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, 9000 Rockville Pike, Bldg 10, Room 12C103A, Bethesda, MD 20892; e-mail: guzel@niaid.nih.gov.