Risk factors of infection-associated mortality in children with lupus nephritis in under-resourced areas

Mortality in children and adolescents with autoimmune inflammatory rheumatic diseases admitted to the pediatric intensive care unit

BACKGROUND

This study aimed to describe the characteristics and outcomes of children and adolescents with autoimmune inflammatory rheumatic diseases (AIIRD) who were admitted to the pediatric intensive care unit (PICU). The accuracy of the Pediatric Risk of Mortality (PRISM) III and Pediatric Index of Mortality (PIM) 3 scores to predict the mortality were investigated.

METHODS

This was a retrospective cohort study. Children and adolescents with AIIRD aged ≤ 18 years who were admitted to the PICU at the largest university-based referral center in Thailand during July 2011 to June 2021 were included.

RESULTS

There were 122 PICU admissions from 74 patients; mean age of 12.0 ± 4.3 years, 74.3% female. Majority of AIIRD were systemic lupus erythematosus (SLE) (83.8%), followed by systemic juvenile idiopathic arthritis (5.4%), juvenile dermatomyositis (JDM) (2.7%) and microscopic polyangiitis (2.7%). The main cause of admission was combined infection and disease flare (29.5%). Pneumonia was the main site of infection. Acinetobacter baumanii was the most common causative agent. Macrophage activation syndrome occurred in 8 (6.5%) admissions. The mortality rate of PICU admissions was 14.8% from 18 deaths; 17 with SLE and 1 with JDM. Mechanical ventilation (aOR 24.07, 95%CI:1.33-434.91, P= 0.031), pneumothorax (aOR 24.08, 95%CI:1.76-328.86, P = 0.017 and thrombocytopenia (aOR 8.34, 95%CI:1.31-53.73, P = 0.025) were associated with mortality. The risk of mortality rate as predicted by the PRISM III score increased with a score ≥ 9. For the PIM 3 score, the risk of mortality increased if the score ≥ 3. The area under the ROC curve for the PRISM III and PIM 3 scores was 0.741 (95%CI: 0.633-0.849), P = 0.001 and 0.804 (95%CI: 0.685-0.924), P < 0.001, respectively. The model calibration using the Hosmer-Lemeshow goodness of fit test demonstrated a chi-square of 4.335, P = 0.826 for PRISM III and 7.987, P = 0.435 for PIM 3.

CONCLUSION

SLE was the main AIIRD that required admission to the PICU. Mechanical ventilation, pneumothorax and thrombocytopenia were associated with mortality in pediatric patients with AIIRD. The PRISM III and PIM 3 scores demonstrated good calibration, while the PIM 3 score provided better discrimination ability in the prediction of mortality for pediatric AIIRD.

Clinical characteristics, disease flares, infections and their impact on kidney outcomes in pediatric lupus nephritis: A cohort study

BACKGROUND

Limited data exists on clinical predictors of kidney outcomes in children with lupus nephritis (LN).

METHODS

Children aged below 18 years with biopsy-proven LN followed from January 2010 to February 2024 in a tertiary-care center were enrolled in order to characterize their clinical presentations, flares, infections, histology and determine Major Adverse Kidney Events (MAKE) (defined as eGFR < 60 mL/min/1.73 m2 and/or death). Data was analysed to identify predictive factors of adverse outcomes.

RESULTS

Seventy-five children (20% boys) with median (IQR) age at diagnosis 11.2 (9,13) years, were studied. Clinical presentations were nephritic syndrome, mixed nephritic-nephrotic features, rapidly progressive glomerulonephritis (RPGN), and nephrotic syndrome in 24 (32%), 15 (20%), 11 (14.7%) and 5 (6.7%) patients, respectively. Proliferative LN was the major (70.6%) histological subtype. In total, 75 kidney flares (21.3% nephritic) with incidence rate (IR) of 0.48 flares per person-year were noted in 31 (41.3%) children. Infections occurred in 32 (42.6%), with IR of 0.58 episodes per person-year. Bacterial pneumonia 14 (22.9%), sepsis 10 (16.3%) and tropical infections 6 (9.8%) were most common. At median last follow-up of 2.3 (1.3, 5.6) years with 85.4% kidney-survival rate, 41 (54.6%), and 21 (28%) were in complete-response (CR), and partial-response (PR), respectively. Proliferative LN and those in PR or NR were at significantly higher risk of kidney flares and infections, regardless of initial induction therapy. RPGN at presentation, non-responders at 6 months and severe kidney flare ever predicted MAKE in 11 (14.6%) children.

CONCLUSIONS

Multiple kidney flares and infections constitute a significant morbidity in LN. RPGN, non-responders and severe kidney flare predict adverse kidney outcomes.

Extracellular traps in peripheral blood mononuclear cell fraction in childhood-onset systemic lupus erythematosus

Although the role of low-density granulocytes (LDGs), neutrophils in the peripheral blood mononuclear cell (PBMC) fraction, and neutrophil extracellular traps (NETs) in assessing lupus disease severity is acknowledged, data specific to childhood-onset lupus remains scarce. This study analyzed 46 patients with childhood-onset systemic lupus erythematosus (82.6% females, mean age 14.5 ± 0.3 years), including 26 cases with normal complement levels and 20 with low complement levels, along with 20 healthy adult volunteers. Key parameters that distinguished healthy volunteers from lupus patients and differentiated between lupus patients with low and normal complement were serum interferon (IFN)-α, serum citrullinated histone 3 (CitH3), and extracellular traps (ETs) in LDGs. However, NETs (assessed by nuclear staining morphology), LDG abundance, and other parameters (such as endotoxemia, cytokines, and double-stranded (ds) DNA) did not show such differentiation. When lipopolysaccharide (LPS) was administered to LDGs in the PBMC fraction, it induced ETs in both low and normal complement groups, indicating the inducible nature of ETs. In adult healthy volunteers, activation by recombinant IFN-α or dsDNA in isolated neutrophils induced LDGs and NETs (identified using immunofluorescent staining for CitH3, myeloperoxidase, and neutrophil elastase) at 45 min and 3 h post-stimulation, respectively. Additionally, approximately half of the LDGs underwent late apoptosis at 3 h post-stimulation, as determined by flow cytometry analysis. Activation by IFN-α or dsDNA in LDGs also led to a more pronounced expression of CD66b, an adhesion molecule, compared to regular-density neutrophils, suggesting higher activity in LDGs. In conclusion, IFN-α and/or dsDNA in serum may transform regular-density neutrophils into LDGs before progressing to NETosis and apoptosis, potentially exacerbating lupus severity through cell death-induced self-antigens. Therefore, LDGs and ETs in LDGs could provide deeper insights into the pathophysiology of childhood-onset lupus.

Hospitalization and Mortality Due to Infection Among Children and Adolescents With Systemic Lupus Erythematosus in the United States

OBJECTIVE

We aimed to determine the frequency and types of infections in hospitalized children with childhood-onset systemic lupus erythematosus (cSLE), and to identify risk factors for intensive care unit (ICU) admission and mortality.

METHODS

We conducted a retrospective study of youth aged 2 to 21 years using International Classification of Diseases (ICD) codes for SLE assigned during admission to a hospital participating in the Pediatric Health Information System, a database of United States children's hospitals, from 2009 to 2021. Generalized linear mixed effects models were used to identify risk factors for ICU admission and mortality among children hospitalized with infection.

RESULTS

We identified 8588 children with cSLE and ≥ 1 hospitalization. Among this cohort, there were 26,269 hospitalizations, of which 13% had codes for infections, a proportion that increased over time (P = 0.04). Bacterial pneumonia was the most common hospitalized infection. In-hospital mortality occurred in 0.4% (n = 103) of cSLE hospitalizations for any indication and 2% of hospitalizations for infection (n = 60). The highest mortality rates occurred with Pneumocystis jirovecii pneumonia (21%) and other fungal infections (21%). Lupus nephritis (LN) and endstage renal disease (ESRD) were associated with increased odds of ICU admission (odds ratio [OR] 1.47 [95% CI 1.2-1.8] and OR 2.40 [95% CI 1.7-3.4]) among children admitted for serious infection. ESRD was associated with higher mortality (OR 2.34 [95% CI 1.1-4.9]).

CONCLUSION

Hospitalizations with ICD codes for infection comprised a small proportion of cSLE admissions but accounted for the majority of mortality. The proportion of hospitalizations for infection increased over time. LN and ESRD were risk factors for poor outcomes.

Neutrophil-to-Lymphocyte Ratio as a Predictor of the Need for Surgical Treatment in Children's Intussusception

OBJECTIVE

 Neutrophil-to-lymphocyte ratio (NLR) is an emerging inflammatory marker in abdominal pathologies. Ileocolic intussusception (ICI) involves a progressive intestinal inflammation, and the effectiveness of nonsurgical treatment (enema) might be related to the inflammation degree, although no previous studies have investigated this relationship. Our aim is to identify predictors of the need for surgical treatment in ICI.

MATERIALS AND METHODS

 A single-center, retrospective, case-control study was performed in children with ICI, who were treated with initial nonsurgical management between 2005 and 2019. Patients were divided in two groups: A (effective enema) and B (need for surgery). Admission demographic and clinical and laboratory data were analyzed. Specificity and sensitivity of the different parameters as predictors of the need for surgical treatment were determined by receiver operating characteristic (ROC) curves.

RESULTS

 A total of 511 patients were included (410: group A; 101: group B), without statistically significant demographic differences. Group B presented significantly higher frequency of vomiting, bloody stools, and longer median time since symptoms onset (24 vs. 8 hours; p < 0.001). Group B presented higher median laboratory inflammatory markers than group A: NLR (6.8 vs. 1.8; p < 0.001), neutrophils (10,148 vs. 7,468; p < 0.001), and C-reactive protein (CRP; 28.2 vs. 4.7; p < 0.001). In ROC curve analysis, NLR had an area under the curve of 0.925, higher than neutrophil count (0.776; p = 0.001), CRP (0.670; p = 0.001), and time since symptoms onset (0.673; p = 0.001). It was estimated a cut-off point of NLR greater than 4.52 (sensitivity: 73.2%; specificity: 94.5%).

CONCLUSION

 High NLR values imply a high degree of bowel inflammation and might anticipate the need for surgical treatment in ICI in children.

LEVEL OF EVIDENCE

 III.

Infections in hospitalized children with newly diagnosed systemic lupus erythematosus in underresourced areas

Background

The administration of induction immunosuppressive therapy to children with newly diagnosed systemic lupus erythematosus (SLE) and concurrent infections can lead to unfavourable outcomes. This study was conducted to describe characteristics of infections occurring before the initiation of immunosuppressants in hospitalized children with newly diagnosed SLE in underresourced areas.

Methods

Medical records of paediatric patients with the diagnosis of SLE, who were admitted to a university-based hospital from 2002 to 2018, were reviewed. Only patients younger than 18 years of age with newly diagnosed SLE were included in the study. The primary outcome was infection before the administration of immunosuppressants. Logistic regression analysis was used to determine factors associated with infection and adjusted odds ratio (OR). The diagnostic accuracy of CRP was assessed.

Results

Infections were confirmed in 52/124 (41.9%) children. Pathogens were identified in 24 (46.2%) patients with bacterial predominance. The most common site was respiratory infections (36.5%). Fever and serosal involvement were more prevalent in patients with infection. Serum CRP levels were significantly higher in patients with infection than in those without infection (median 5.5 mg/L (interquartile range (IQR) 3.6–76.3 mg/L) vs. 3.5 mg/L (IQR 3.0–3.6 mg/L), p = 0.004). When a positive CRP level of >5 mg/L was used, positive CRP was found with a higher prevalence in patients with infection and was independently associated with infection (adjusted odds ratio (OR) = 28.6, 95% confidence interval (CI) 2.3–350.6; p = 0.009). Patients with infection had a longer hospital stay than patients without infection (median 20 days (IQR 13–25 days) vs. 15 days (IQR 9–24 days), p = 0.04). Sensitivity, specificity, positive predictive value and negative predictive value with 95% CI of CRP >5 mg/L were 62.5% (35.4–84.8%), 88.9% (65.3–98.6%), 80.3% (51.0–94.1%) and 76.6% (63.1–86.3%), respectively.

Conclusions

Infections were common among hospitalized children with newly diagnosed SLE. Children with infections had a prolonged course of hospitalization. Positive CRP was associated with a predisposition towards infection. However, the diagnostic accuracy of CRP requires further validation in a larger study.