Authors: Guanhua Li, Xiaojing Liu, Jiakang Zhao
Categories: Research Article, Systemic lupus erythematosus, Infection, Autoimmune disease
Source: Journal of Community Hospital Internal Medicine Perspectives
Authors: Guanhua Li, Xiaojing Liu, Jiakang Zhao
To determine risk factors, clinical and microbiological characteristics of infections in a single-center systemic lupus erythematosus (SLE) cohort.
All hospital patients in The First Affiliated Hospital of Zhengzhou University from 2019 to 2021 who meet ≥4 ACR-97 SLE criteria were identified. Patients with infection and without infection were included with a ratio of 2.
687 SLE patients were identified and 224 patients with infection and 448 patients without infection were included. The most common microorganisms in the infection group were Escherichia coli, Herpes zoster and Aspergillus, respectively. In the cox regression analysis, nephritis (OR 1.785; 95% CI: 1.156–2.756), diabetes(OR 6.507; 95% CI: 2.692–15.725), CD4+ T cell (OR 1.003; 95% CI: 1.002–1.004), prednisone>10 mg/day (OR 1.879; 95% CI: 1.112–3.175), immunosuppressants (OR 0.465; 95% CI: 0.310–0.697) and SLEDAI score (OR 0.866; 95% CI: 0.837–0.896) were risk factors associated with infection.
Bacteria was the most common infection in SLE patients with the respiratory tract being the most common site. Nephritis, diabetes, SLEDAI score were associated with infection. Monitoring CD4+ T cell can predict the infection incidence. Prednisone (>10 mg/day) and immunosuppressants increase the occurrence of infection.
Systemic lupus erythematosus is an autoimmune disease that can impact many organs. A complex interaction of environment, genetics and hormones lead to resistant treatment and high mortality.1 In the past few decades, the clinical treatment of SLE has made great progress, but there still remains some challenges.2 Infection accounts for 45% of the morbidity and 11–23% hospitalizations in SLE patients. Besides, Lupus flares and recurrent complications are the main cause of poor prognosis. The flare and repeated hospital admission bring a heavy economic burden on society and for the family.3–6 Therefore, it is a great challenge for physicians to notice risk factors of infection and protect SLE patients from infection. Common SLE treatments like corticosteroids and other immunosuppressants increase the risk for infection.7–9
There is very little data from Middle China in the literature and we aim to report the incidence, microbiologic, clinical characteristics and the risk factors for infection in SLE patients from our center in Middle China.
SLE patients admitted in The First Affiliated Hospital of Zhengzhou University from 2019 to 2021 who met at least 4 ACR-97 SLE criteria were identified. Pregnant patients and patients with incomplete data were excluded. A total of 224 patients with infection were included. Then 448 SLE patients without infection were matched according to sex and age at a rate of 2 as controls. We collect demographics, comorbidities, SLE Disease Activity Index (SLEDAI, a validated index that assesses the severity and change of lupus disease activity), laboratory data, medical treatment at the time of infection. Comorbidities consist of Lupus nephritis, diabetes, dialysis, anemia and thrombocytopenia. Laboratory data contained hemoglobin, blood platelet, C3, C4, IgG, CD4+ T cell and CD8+ T cell. Medical treatment mainly included azathioprine, leflunomide, methotrexate, mycophenolate mofetil (MMF), tacrolimus, cyclophosphamide, hydroxychloroquine and daily dose of prednisone. Glucocorticoid was converted as 1 mg prednisone = 0.15 mg dexamethasone = 0.8 mg methylprednisolone. All the above drugs had been used continuously for at least three months before infection. The study was approved by the ethics committee of The First Affiliated Hospital of Zhengzhou University.
Infections were defined on the basis of positive cultures, next generation sequence, elevated temperature (>38 °C), the abnormality of white blood cell (<4 × 10^9^/L or > 10 × 10^9^/L), the high level of erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP), and the response to antibiotic therapy.
Numerical variable are presented as mean ± standard deviation and categorical variable are presented as frequencies and percentages. Chi-square test was performed for qualitative-independent variables and Student’s t test for quantitative-independent variables. A logistic regression was conducted to assess risk factors. The IBM-SPSS26 was applied for all statistical analysis. Significance was defined as p < 0.05.
We included 224 SLE patients with infection(females 195 and males 29), with 448 SLE patients( females 394 and males 54) as controls without infection at a rate of 2 according to age and sex (Table 1). The mean SLEDAI score was 12.33 ± 7.44 and 6.88 ± 5.10 in controls (p < 0.001). Refer to baseline comorbidity, the infection group has higher incidence than the controls, in the infection group 73 (32.6%) was lupus nephritis and 206 (46.0%) in the controls (p < 0.001), 26 (11.6%) was diabetes and 14 (3.1%) in the control (p < 0.001), 20 (8.9%) suffered dialysis and 10 (2.2%) in the control (p < 0.001), 204 (91.1%) were diagnosed with anemia and 376 (83.9%) in the control (p < 0.001), 70 (31.2%) were diagnosed with thrombocytopenia and 73 (16.3%) in the controls (p < 0.001). At the same time, laboratory data showed the rate of low C3 in infection group was 140(62.5%) and 295 (65.8%) in the controls (p = 0.392), low C4 was 95 (42.4%) and 213 (47.5%) in the controls (p = 0.208), IgG level was 11.84 ± 9.16 g/L vs 13.01 ± 6.24 g/L (p = 0.18), CD4+ T cell (cells/UL) was 268.78 ± 210.10 vs 395.07 ± 310.61(p < 0.001), CD8+ T cell (cells/UL) was 317.69 ± 292.03 vs 372.43 ± 298.94 (p = 0.937). Regarding the treatment prior to infections, the use of MMF, prednisone>10 mg/day, tacrolimus had significant difference, which the rate was 58 (34.9%) vs 153 (34.2%) (p = 0.030), 187 (83.5%) vs 335 (74.8%) (p < 0.001) and 32 (16.7%) vs 38 (8.5%) (p = 0.020), respectively. The use of azathioprine, leflunomide, methotrexate, cyclosporine, cyclophosphamide, hydroxychloroquine had no difference (Table 1). Table 2 shows the ACR97 SLE criteria distribution. The presence of renal disorders, neurologic disorders, immunologic disorders are more common in the SLE patients with infection. The presence of malar rash, discoid rash, photosensitivity, oral ulcers, arthritis, serositis, hematologic disorders had no significant difference (Table 2). In conclusion, the SLE patients with infections had more abnormal clinical data, baseline comorbidity and higher SLEDAI score.
Among the 224 SLE patients with infection, 147(65.63%) were bacterial infection with 92(62.59%) respiratory tract, 19(12.93%) urinary tract, 17(11.56%) bacteremia/sepsis, 10(6.80%) skin and mucous membranes, 2(1.36%) peritonitis. Meanwhile, incidence of catheter-related, suppurative appendicitis, lymphadenopathy, meningitis, pericarditis, lumbar vertebra, and pancreatic infection was 1(0.68%). As to 22(9.82%)viral infection, 13(59.09%) was respiratory tract, 6(27.27%) skin and 3(13.64) central nervous system. Incidence of fungal infection was 61(27.23%), including 57(93.44%) respiratory tract, 2(3.28%) upper gastrointestinal tract and mouth and throat (Table 3).
Compared with SLE patients without infections, higher SLEDAI score, high percentage of lupus nephritis, diabetes, dialysis, anemia, thrombocytopenia, usage of tacrolimus, MMF, prednisone> 10 mg/day and low level of CD4+ T cell (cells/UL) were more common in SLE patients with infections. We further analyzed risk factors by logistic regression analysis. The results showed that nephritis (OR = 1.785; 95%CI:1.156–2.756), diabetes (OR = 6.507; 95%CI:2.692–15.725), CD4 T cell (OR = 1.003; 95%CI:1.002–1.004), prednisone>10 mg/day (OR = 1.879; 95%CI:1.112–3.175), immunosuppressants (OR = 0.465; 95%CI:0.310–0.697) and SLEDAI score (OR = 0.866; 95%CI:0.837–0.896) were independent risk factors for infections in SLE patient (Table 4).
In this study, we demonstrated the clinical and microbiological characteristics of SLE patients with infection. We further analyzed the risk factors in SLE patients with infections, mainly including higher SLEDAI score, lupus nephritis, diabetes, usage of immunosuppressants, prednisone>10 mg/day and low level of CD4+ T cell.
Compared with the SLE patients without infection, the group with infection had higher mean SLEDAI score, which is consistent with a previous study.10 Previous studies have showed respiratory tract, urinary infection and bacteriaemia are most common in SLE patients.11 Our data supports that of previous studies that the respiratory tract is the most common infection site in the bacterial, viral and fugal groups (Table 3). Apart from the common infection sites, there are some rate infection site like catheter-related infection, lumbar vertebra, and meningitis. In this study, microorganisms were detected in 90 (40.18%) out of 224 patients, next generation sequence(NGS) improved the diagnosis rate. A total of 27 pathogen strains were detected. Among the 27 pathogen strains, most were detected by microbial cultures, however there are still some limitations.12 For example, low concentration of pathogen may need a long time to be detected in the culture and not all the microbes grow in the standard conditions. With the NGS in clinical practice, some culture-negative microorganisms were identified by NGS eventually.13 As the rapid and accurate confirmation of pathogenic agents, it is of great value to guide the diagnosis and treatment. NGS may be an effective and essential supplementary method for the diagnosis of infection.14 In line with previous studies, bacterial was the most common source of infection (65.6%), especially Escherichia coli, Salmonella SPP and Streptococcus pneumoniae. Different from other studies, Acinetobacter baumannii was not the most common bacteria. In the viral infections, Epstein–Barr virus was most common in the respiratory tract and Herpes zoster was most popular among skin sites, which was in accordance with previous studies. Fingal infection were rather common and the most common was Candida. Aspergillus was second to the Candida. Different with other studies,15 cryptococcus was not isolated in our study. This may due to the regional differences or sample selections.
The advantage and disadvantage of glucocorticoids and immunosuppressive on the patients with SLE have been reported by many studies.16–18 Even more, the ever use of glucocorticoid may increases the rate of infection in the future.19 Our data confirmed this issue. Patients receiving prednisone> 10 mg/day increase the susceptibility to infection. By multivariate logistic regression, our study showed that prednisone>10 mg/day was an independent risk factor for infection in SLE patients. Despite the fact that many studies found that cyclophosphamide increase the infection possibility and act as an independent risk factor for infection in SLE patients.20 In our study, as to the immunosuppressive, only MMF and Tacrolimus increase the infection possibility, while cyclophosphamide had no difference. In the multivariate logistic regression the use of immunosuppressive was an independent risk factor, but the single type of any immunosuppressive showed no meaning. This may be explained by the popular use of MMF and tacrolimus in the clinical, bias of patients selected and the limited patients number. More studies are needed to confirm the results.
A complex interaction of adaptive immune cell and innate immune cell leads to unbalanced immune system, resulting in abnormal antibody production and inflammation. Of these immune cells, T cells contribute significantly to the pathogenesis of SLE and are meaning to maintain immune homeostasis. CD4+ T cell usually protect against infections and lower level of CD4+ T lymphocytes indicated defective host immune system.1 Most of patients enrolled in our study showed a decreased CD4+ T lymphocytes and low level of CD4+ T lymphocytes is associate with more infection and act as an independent risk factor in SLE patients. In line with other researchers, our research showed decreased CD4+ T lymphocytes increase the risk of infection in SLE patients.21 Therefore, it is necessary to monitoring the number of CD4+ T lymphocytes in SLE patients. Once the decreased CD4+ T lymphocytes is detected, Chinese drugs such like Kangai can be applied to the patients to balance immunity.22 Another typical characteristic in SLE patients is low C3 and C4, the abnormal complement not only impaired clearance of variety immune complexes and regulation of cytokine, but also give rise to recurrent infection and organ damage.23 Unlike these studies, our data found there was no difference between SLE patients with and without infections. Considering the limitations of clinical studies, more basic research should be conducted to clarify the issue. Baseline comorbidity such as nephritis and diabetes were associated with infection. As we all know, the high level of glucose is a good environment for bacteria and disturb the immune system.24
In conclusion, bacteria is the most common infection in SLE patients, resulting from higher SLEDAI score, lupus nephritis, diabetes, anemia, thrombocytopenia, usage of immunosuppressants, prednisone>10 mg/day and low level of CD4+ T cell. Infection in SLE patients is associated with lupus nephritis, diabetes, usage of immunosuppressants, prednisone>10 mg/day and low level of CD4+ T cell. Monitoring CD4+ T cell and avoiding glucocorticoid can reduce the occurrence of infection.
As far as we know, this is the first study demonstrating the microorganism characteristics of SLE patients with and without infection in middle China. Moreover, we included the NGS data to improve the diagnosis rate. Nevertheless, this study has some limitations. First, data was collected from only one hospital and just the inpatient were included. As the hospital was the provincial hospital, most patients are serious ill, this possible improved the infection rate. Second, important information regarding medicine (glucocorticoids and immunosuppressive) duration and disease course were not recorded. This may result the bias in final analysis. Third, we only collected data from adult apartments, missing most early-onset SLE patients, the data can only serve as a reference for adult SLE patients.