Serious infections and tuberculosis adversely impact outcomes of juvenile onset systemic lupus erythematosus in India

Abstract

Background

Infections are a major cause of morbidity and mortality in juvenile systemic lupus erythematosus (SLE). We assessed the incidence and risk factors for major infections in juvenile SLE.

Methods

A retrospective review of 225 patients of juvenile SLE (ACR 1997 criteria) with age <18 years visiting the rheumatology clinic at a single centre between 2000 to 2020 was done from case records and the hospital electronic health records. Serious infection was defined as the need for hospitalization, or infection resulting in disability or death. Cox regression was used to determine factors associated with a serious infection and the effect of serious infection on overall survival.

Results

We reviewed 225 children (197 girls, mean age 13.89 ± 3.42 years) with a cumulative follow up of 1153.45 person-years. Eighty serious infections occurred in 63 (28% of the cohort) children at a rate of 69.35 serious infections per 1000 person-years. A second serious infection occurred in 12 children and 5 of them developed three infections.

Among the cases with known etiology (78.75% of cases), bacterial infections were most common (N = 33) including S. Aureus (11), E. Coli (7), K. Pneumoniae (3), E. Fecalis (3), S. Pneumoniae (2), Acinetobacter spp. (2), Citrobacter (2), Salmonella (2) and P. Aeruginosa (1). Twenty six (32.5%) opportunistic infections occurred: Mycobacterium tuberculosis (18), Cytomegalovirus (3), disseminated Herpes zoster (4) and invasive candidiasis (1) with 15 (83.3%) of the tuberculosis cases being extrapulmonary. On multivariate analysis, fever (HR 8.51, 1.17–61.44), gastrointestinal involvement (HR 4.73, 1.13–19.94), current steroid dose (HR 1.36,1.14–1.62), average cumulative steroid dose per year (HR 1.004, 1.002–1.005) and cyclophosphamide (HR 2.22, 1.11–4.46) were associated with serious infection.

Hospitalization rates were significantly higher in those with any serious infection (Rate-ratio 2.79, 1.81–3.77) as was damage accrual (SLICC damage index 1.04 vs 0.22). Serious infection-free survival at 1 year and 5 years was 84% (79.1–89.2) and 72% (65.4–79.2). There were 19 deaths with infection attributable mortality in 10 (52.6%). Serious infection predisposed to higher overall mortality with recurrent infections conferring a hazard ratio of 36.02 (8.07–160.62).

Conclusion

Serious infections are a major cause of mortality and damage in SLE. Constitutional symptoms, gastrointestinal involvement, current and cumulative steroid dose and cyclophosphamide predict serious infections. TB prophylaxis in patients with SLE should be considered in endemic areas, especially when using high-dose steroid therapy.