Pediatric lupus nephritis - An evil cousin of its adult counterpart: A single-center based experience from a tertiary care hospital of Eastern India

Lupus nephritis or not? A simple and clinically friendly machine learning pipeline to help diagnosis of lupus nephritis

OBJECTIVE

Diagnosis of lupus nephritis (LN) is a complex process, which usually requires renal biopsy. We aim to establish a machine learning pipeline to help diagnosis of LN.

METHODS

A cohort of 681 systemic lupus erythematosus (SLE) patients without LN and 786 SLE patients with LN was established, and a total of 95 clinical, laboratory data and 17 meteorological indicators were collected. After tenfold cross-validation, the patients were divided into training set and test set. The features selected by collective feature selection method of mutual information (MI) and multisurf were used to construct the models of logistic regression, decision tree, random forest, naive Bayes, support vector machine (SVM), light gradient boosting (LGB), extreme gradient boosting (XGB), and artificial neural network (ANN), the models were compared and verified in post-analysis.

RESULTS

Collective feature selection method screens out antistreptolysin (ASO), retinol binding protein (RBP), lupus anticoagulant 1 (LA1), LA2, proteinuria and other features, and the hyperparameter optimized XGB (ROC: AUC = 0.995; PRC: AUC = 1.000, APS = 1.000; balance accuracy: 0.990) has the best performance, followed by LGB (ROC: AUC = 0.992; PRC: AUC = 0.997, APS = 0.977; balance accuracy: 0.957). The worst performance is naive Bayes model (ROC: AUC = 0.799; PRC: AUC = 0.822, APS = 0.823; balance accuracy: 0.693). In the composite feature importance bar plots, ASO, RF, Up/Ucr, and other features play important roles in LN.

CONCLUSION

We developed and validated a new and simple machine learning pathway for diagnosis of LN, especially the XGB model based on ASO, LA1, LA2, proteinuria, and other features screened out by collective feature selection.

Analysis of Clinicopathological Characteristics and Its Correlation With the Prognosis of Pediatric Lupus Nephritis: A Tertiary Care Center Experience

Aim

To study the various pathological patterns of pediatric lupus nephritis (LN) by renal biopsies and to correlate the histopathological data with the clinical and biochemical outcomes.

Methods

This is a retrospective study in children between 1 month and 18 years of age with renal biopsy-proven lupus nephritis, conducted between January 2015 and December 2019. Various pathological and clinical parameters were compared between the groups with lupus nephritis activity and those without activity.

Results

Of 38 biopsy-proven lupus nephritis cases, 30 (78.9%) were in the adolescent age group, and the female gender was predominantly affected (n=30; 78.9%). Class IV proliferative lupus nephritis (n=17, 44.7%) was the most common biopsy finding, and the activity score for endocapillary hypercellularity, neutrophil infiltration, fibrinoid necrosis, hyaline deposits, and interstitial inflammation was significantly high in classes III and IV. Overall, attaining remission was less, and the risk of progression of chronic kidney disease (CKD) was higher in class IV (n=3; 7.8%). Mortality was reported in 1 out of 38 (2.6%) children.

Conclusion

Light microscopy and immunofluorescence studies play an important role in defining the extent of renal damage in the form of activity and chronicity indices, which are the key factors in the decision-making of lupus nephritis treatment. The prognostic relevance of the histological scoring has been evaluated, and it is evident that the activity index and chronicity index go a long way in therapeutic intervention.