Childhood-onset systemic lupus erythematosus in China, 2016-21: a nationwide study.

BACKGROUND

Childhood-onset systemic lupus erythematosus (SLE) is a complex autoimmune disorder that can lead to serious organ damage. Despite the prevalence of SLE among children in Asian countries, treatment guidelines, prognosis, and clinical decision making for children with SLE are limited by gaps in region-specific epidemiological data. The aim of this study was to analyse epidemiological characteristics of childhood-onset SLE and associated organ involvement and in-hospital mortality in China.

METHODS

In this nationwide study, we searched standardised hospital discharge records submitted to the Hospital Quality Monitoring System (HQMS) between Jan 1, 2016, and Dec 31, 2021, for patients with a diagnosis of childhood-onset SLE based on the 2019 American College of Rheumatology or 2012 Systemic Lupus International Collaborating Clinics classification criteria. We selected records for patients aged 18 years and younger containing relevant ICD 10th revision diagnostic codes (specifically M32) among discharge diagnostic codes. We excluded records for patients younger than 5 years, whose SLE diagnosis was presumed to be monogenic lupus, and for patients with overlap syndromes or unidentified sex. Date of diagnosis (equal to the first hospital discharge date), age, organ involvement, intensive care unit (ICU) treatment, and in-hospital mortality were extracted from the records. Incidence rates for 2017, 2018, 2019, 2020, and 2021 were identified with five washout periods ranging from 12 months (Jan 1-Dec 31, 2016) to 60 months (Jan 1, 2016-Dec 31, 2020). Data were stratified by sex, age relative to puberty onset, organ involvement and concurrent infection at time of diagnosis, human development index of region of residence, hospital level, and hospital type. Incidence trends by sex, age relative to puberty onset, and year were derived by joinpoint regression analysis, with 95% CIs calculated by the Poisson exact method. Major organ involvement was assessed according to definitions in the British Isles Lupus Assessment Group 2004 disease activity index. Outcomes of ICU admission after first diagnosis and in-hospital death after ICU admission were analysed in Cox proportional hazards models, with p values and 95% CIs calculated with the parametric method.

FINDINGS

Between Jan 1, 2016, and Dec 31, 2021, the HQMS received 134 956 hospital discharge records containing the M32 discharge diagnostic code for patients aged 18 years or younger. 6286 records were excluded, leaving 128 670 records representing 54 338 patients aged 5-18 years; of these, 43 756 patients (36 153 girls and 7603 boys) received their childhood-onset SLE diagnosis on or after Jan 1, 2017. Between Jan 1, 2017, and Dec 31, 2021, the SLE incidence rate was 3·97 (95% CI 3·93-4·01) per 100 000 person-years, with a declining trend during the 5-year period. Joinpoint analysis showed sex-dependent and age-dependent incidence patterns. After a relatively stable incidence rate among prepubertal children aged 5-7 years, the incidence rate increased among peripubertal children aged 8-12 years by 0·92 cases per 100 000 person-years with each 1-year increase in age (p<0·0001). Among peripubertal children aged 8-12 years, girls showed the largest change in incidence rate, with an increase of 1·64 per 100 000 person-years with each 1-year increase in age (p<0·0001), compared with 0·40 per 100 000 person-years with each 1-year increase in age among boys (p=0·013). The organ systems most affected in patients with childhood-onset SLE were the kidneys (56·8%) and the haematological system (27·8%). Among the 2471 patients admitted to the ICU, 213 (9%) of whom died in ICU, the three organ-related risk factors at initial diagnosis that showed greatest association with progression to critical illness were cardiovascular involvement (adjusted hazard ratio 2·50 [95% CI 2·18-2·87]; p<0·0001), neuropsychiatric SLE (2·10 [1·87-2·37]; p<0·0001), and serositis (2·03 [1·78-2·30]; p<0·0001). Other prominent risk factors for progression to critical illness were concurrent infections with Epstein-Barr virus (1·52 [1·16-1·98]; p=0·0020) or fungi (1·49 [1·22-1·83]; p=0·0001). In total, 396 (0·7%) of 54 338 patients with childhood-onset SLE died in hospital; the most common causes of death were pneumonia (146 [37%]), multi-organ dysfunction syndrome (99 [25%]), and renal failure (75 [19%]). Risk of in-hospital mortality was highest among pubertal children (hazard ratio 2·16 [95% 1·14-4·09]) compared with prepubertal children, and risk of ICU admission was highest among prepubertal children (2·95 [2·16-4·03]) compared with postpubertal children.

INTERPRETATION

These nationwide data on the epidemiology of childhood-onset SLE in the Chinese paediatric population show for the first time a declining trend in incidence rates, rapid rise in puberty-onset rates, and the distinct involvement of vital organs from disease onset to mortality in China. They underscore the complexity of childhood-onset SLE pathogenesis and emphasise the imperative for stratified precision treatment, informed interventions, and health-care planning for childhood-onset SLE.

FUNDING

National Key Research & Development Program of China.