Biopsy or Biomarker? Children With Minimal Change Disease Have a Distinct Profile of Urinary Epidermal Growth Factor

Machine learning-based diagnostic prediction of minimal change disease: model development study

Minimal change disease (MCD) is a common cause of nephrotic syndrome. Due to its rapid progression, early detection is essential; however, definitive diagnosis requires invasive kidney biopsy. This study aims to develop non-invasive predictive models for diagnosing MCD by machine learning. We retrospectively collected data on demographic characteristics, blood tests, and urine tests from patients with nephrotic syndrome who underwent kidney biopsy. We applied four machine learning algorithms-TabPFN, LightGBM, Random Forest, and Artificial Neural Network-and logistic regression. We compared their performance using stratified 5-repeated 5-fold cross-validation for the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). Variable importance was evaluated using the SHapley Additive exPlanations (SHAP) method. A total of 248 patients were included, with 82 cases (33%) were diagnosed with MCD. TabPFN demonstrated the best performance with an AUROC of 0.915 (95% CI 0.896-0.932) and an AUPRC of 0.840 (95% CI 0.807-0.872). The SHAP methods identified C3, total cholesterol, and urine red blood cells as key predictors for TabPFN, consistent with previous reports. Machine learning models could be valuable non-invasive diagnostic tools for MCD.

Epidemiology and clinicopathological characteristics of native kidney disease in children in Flanders, Belgium

BACKGROUND

The Flemish Collaborative Glomerulonephritis Group (FCGG) registry is a population-based kidney biopsy registry that has been including all native kidney biopsies performed in children in Flanders (Belgium), since 2017.

METHODS

From 2017 to 2020, 148 pediatric (< 18 years) native kidney biopsies were included. Each biopsy received a histopathological and final nephrological diagnosis, and concordance between both was assessed. Disease chronicity, summarized by the Mayo Clinic Chronicity Score, was determined on 122 biopsies with > 5 glomeruli.

RESULTS

Kidney biopsy rate was high (29.0 biopsies per million children per year), median age was 10.0 years (IQR 5.8-14.7), and boys predominated (56.1% males). A total of 140 biopsies (94.6%) showed a representative pathology result. Glomerular disease was most prevalent, with IgA nephropathy/IgA vasculitis (43 biopsies, 29.1%) and minimal change disease (MCD) (29 biopsies, 19.6%) being the overall most frequent diagnoses. In general, diagnostic concordance was high (80.7%). In Alport syndrome and focal segmental glomerulosclerosis (FSGS), concordance was lower, as the nephrological diagnosis was often determined by results of genetic analysis. Nephrotic syndrome was the most frequent indication for kidney biopsy (31.8%) and was mainly caused by MCD and FSGS. The degree of disease chronicity on kidney biopsies was generally low, although 27.3% of biopsies with a diagnosis of FSGS showed moderate-to-severe chronic damage.

CONCLUSIONS

The presented epidemiological findings validate data from previous European registry studies and may inspire kidney biopsy registries worldwide to implement novel features such as clinicopathological concordance and chronicity grading. A higher resolution version of the Graphical abstract is available as Supplementary information.