Autoimmunity in sickle cell disease: Analysis of a large cohort of adult patients.

Sickle cell disease (SCD) is a genetic hemoglobinopathy leading to chronic hemolysis punctuated by vaso-occlusive episodes and responsible for chronic organ damages.1 While the association between SCD and some autoimmune diseases (AIDs) has been reported, the diagnosis of AIDs in this context is often described as difficult and delayed, due to similarities in clinical manifestations of both conditions.2, 3 Biologically, some authors have reported a higher prevalence of circulating autoantibodies in patients with SCD than in the general population, making the diagnosis of AID more complex.4, 5 Moreover, the association between SCD and AIDs seems to result in a significant morbidity and mortality. However, these observations most often come from isolated case reports and need better characterization. Here, we conducted an observational study with retrospective data analysis of medical files from all consecutive patients aged 18 years or older and followed for SCD between January 2012 and July 2020. We analyzed the prevalence of biological autoimmunity and the occurrence of an established diagnosis of systemic immunological diseases (according to international criteria). We used the SCD clinical severity score used by Michel et al.3 to compare each patient to himself before and after the diagnosis of AID. The treatments and course of both conditions were analyzed. We included 304 SCD patients, mainly women (54.6%) with homozygous SCD (68.4%) and a mean age of 33.3 ± 11.4 years. Routine testing for markers of autoimmunity was performed during the follow-up of 266 of these patients and was positive in 154/266 (57.9%), mainly consisting in ANAs (95.5%) – mostly without specificity (74.7%), sometimes associated with anti-liver antibodies (27.9%) and less frequently anti-CCP (9.7%) and a rheumatoid factor (8.4%). We did not notice any significant difference in terms of sex ratio, mean age, genotype distribution, overall SCD severity score or treatments received whether the patients had biological autoimmunity or not (Table 1). However, the prevalence of G6PD deficiency was higher in the group without biological autoimmunity. Among the 154 patients with biological autoimmunity, only 15 patients presented an overt diagnosis of systemic immunological disease, representing 5.6% (15/266) of the tested cohort and 9.7% (15/154) patients with biological autoimmunity. Eight patients had rheumatoid arthritis (RA), one had lupus associated with myasthenia, one had amyopathic anti-synthetase syndrome, one had autoimmune hepatitis, one had myelitis consistent with neuromyelitis optica, two had sarcoidosis, and one had Crohn's disease. Most patients were female (66.7%) and had SS genotype (66.7%). There were statistically more patients with a history of splenectomy (26.7% vs. 5.0%, p = .01) and renal impairment (53.3% vs. 24.5%, p = .03) in the group with systemic diseases (Table 1). Conversely, patients with isolated biological autoimmunity had a higher prevalence of cerebrovascular abnormalities and retinopathy. The median follow-up after diagnosis of systemic disease was 7 (1–18) years. The management of the systemic disease included systemic corticosteroids (n = 12 [80%]) or immunosuppressants (n = 13 [86.7%]), 10 of which were combined with corticosteroids. Six patients (40%) also received at least one biologic therapy (TNF inhibitors [n = 4], rituximab [n = 2], tocilizumab [n = 1], and ustekinumab [n = 1]). Five patients (41.6%) had uncontrolled AID activity on immunosuppressant. Two patients (16.7%) died during follow-up from complications attributed to SCD but with concomitant active AID. Regarding the course of SCD, 7 of 13 patients on long-term immunosuppressive therapy experienced an exacerbation of SCD (54%), defined as an increase in the severity score of at least 2 points after the diagnosis of systemic disease. On average, the SCD severity score in these patients increased from 2.57 ± 3.26 before the diagnosis of systemic disease to 9 ± 8.27 after a median follow-up of 7 years. Immunosuppression was also associated with an increase in the number of infections requiring hospital management in eight patients (53%). We also studied the characteristics of patients with biological anti-liver autoimmunity. Forty-three patients had anti-liver autoimmunity with positive anti-smooth muscle antibody (ASMA) (53.5%) and anti-LKM1 (41.9%) and only one patient was diagnosed with type I autoimmune hepatitis (AIH). These 43 patients with anti-liver autoimmunity had likely homozygous SCD (86% vs. 69.4%), a higher prevalence of hemochromatosis (32.6% vs. 12.6%) and a significantly higher mean SCD severity score (7.84 ± 6.55 vs. 5.46 ± 5.29) than patients with another type of autoimmunity (Table 1). There were also more frequent background treatments (60% of patients on hydroxyurea and 51% on red cell exchange). In this study, we found a particularly high prevalence of biological autoimmunity among SCD patients, since 58% of the tested sera were positive. This is a higher prevalence than previously reported in the literature, ranging from 10% to 54%.4-6 Some hypotheses to explain this include the role of systemic inflammation, VOC and chronic antigenic stimulation. However, we did not find evidence of a higher disease activity in patients with biological autoimmunity, as previously reported.4, 5 No specific SCD features correlated with the risk of established AID, except for a past history of splenectomy, possibly leading to decreased clearance of immune complexes following splenectomy, enhancing the development of AID. We found a particularly high prevalence of anti-liver autoimmunity, 16% of the autoimmune tests performed being positive for at least one anti-liver antibody. Interestingly, the presence of these antibodies seems to be associated with a more severe SCD phenotype and most often a homozygous genotype. Some authors have suggested that chronic hemolysis could be a factor in the induction of ASMA4 by releasing intracellular contractile proteins. Similarly, one can hypothesize that the recurrence of liver damage (due to pigment gallstones, iron overload, intrahepatic VOC, infections) could lead to the release of autoantigens. As previously described,2, 3 the evolution of AIDs on immunosuppressive treatment was mostly unfavorable since 41.6% of patients had persistent activity and 16.7% died during follow-up. Also, the course of SCD was also unfavorable with more than half the patients presenting a flare after the diagnosis of AID. One hypothesis is that during active AID, the production of pro-inflammatory cytokines (IL-1β, IL-6, TNFα) can stimulate the vascular endothelium, induce erythrocyte adhesion and lead to the occurrence of VOC2. The other hypothesis is an exacerbation related to the immunosuppressive therapy, particularly corticosteroids. There are few data in the literature concerning the use of biological drugs in this context, but the available observations are encouraging with the vast majority having good control of their AID and no worsening of their SCD or infectious episodes.2, 3 In conclusion, autoimmunity seems more frequent in SCD than in the general population. The coexistence of AIDs and SCD appears to be associated with both a severe phenotype of AID and worsening of SCD. However, these observations are limited by the retrospective nature of our study and need to be verified on a larger scale. The data that support the findings of this study are available from the corresponding author upon reasonable request. Data S1. Supporting information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.