The application of deep learning for automated segmentation (delineation of boundaries) of histologic primitives (structures) from whole slide images can facilitate the establishment of novel ...protocols for kidney biopsy assessment. Here, we developed and validated deep learning networks for the segmentation of histologic structures on kidney biopsies and nephrectomies. For development, we examined 125 biopsies for Minimal Change Disease collected across 29 NEPTUNE enrolling centers along with 459 whole slide images stained with Hematoxylin & Eosin (125), Periodic Acid Schiff (125), Silver (102), and Trichrome (107) divided into training, validation and testing sets (ratio 6:1:3). Histologic structures were manually segmented (30048 total annotations) by five nephropathologists. Twenty deep learning models were trained with optimal digital magnification across the structures and stains. Periodic Acid Schiff-stained whole slide images yielded the best concordance between pathologists and deep learning segmentation across all structures (F-scores: 0.93 for glomerular tufts, 0.94 for glomerular tuft plus Bowman’s capsule, 0.91 for proximal tubules, 0.93 for distal tubular segments, 0.81 for peritubular capillaries, and 0.85 for arteries and afferent arterioles). Optimal digital magnifications were 5X for glomerular tuft/tuft plus Bowman’s capsule, 10X for proximal/distal tubule, arteries and afferent arterioles, and 40X for peritubular capillaries. Silver stained whole slide images yielded the worst deep learning performance. Thus, this largest study to date adapted deep learning for the segmentation of kidney histologic structures across multiple stains and pathology laboratories. All data used for training and testing and a detailed online tutorial will be publicly available.
Display omitted
Heterogeneity in disease course and treatment response among patients with MCD/FSGS necessitates a granular evaluation of kidney tissue features. This study aimed to identify histologic and ...ultrastructural descriptors of structural changes most predictive of clinical outcomes in the Nephrotic Syndrome Study Network (NEPTUNE).
Forty-eight histologic (37 glomerular, 9 tubulointerstitial, 2 vascular) and 20 ultrastructural descriptors were quantified by applying the NEPTUNE Digital Pathology Scoring System to NEPTUNE kidney biopsies. Outcomes included time from biopsy to disease progression, first complete remission of proteinuria, and treatment response. Relative importance of pathology and clinical predictors was obtained from random forest models, and predictive discrimination was assessed.
Among 224 participants (34% Black, 24% Hispanic), model performance was excellent, with predictive discrimination of 0.9 for disease progression, 0.85 for complete remission, and 0.81 for treatment response. The most predictive descriptors of outcomes included both conventional-
, global sclerosis or segmental sclerosis and interstitial fibrosis/tubular atrophy-and novel features, including adhesion, interstitial foam cells, deflation, periglomerular fibrosis, mononuclear white blood cells, endothelial cell abnormalities, microvillous transformation, and acute tubular injury.
The most predictive descriptors of clinical outcomes among MCD/FSGS patients reflected structural changes in multiple renal compartments. Reporting these descriptors should be standardized to guide prognostication of proteinuric glomerular diseases.
The current classification system for focal segmental glomerulosclerosis (FSGS) and minimal change disease (MCD) does not fully capture the complex structural changes in kidney biopsies nor the ...clinical and molecular heterogeneity of these diseases.
Prospective observational cohort study.
221 MCD and FSGS patients enrolled in the Nephrotic Syndrome Study Network (NEPTUNE).
The NEPTUNE Digital Pathology Scoring System (NDPSS) was applied to generate scores for 37 glomerular descriptors.
Time from biopsy to complete proteinuria remission, time from biopsy to kidney disease progression (40% estimated glomerular filtration rate eGFR decline or kidney failure), and eGFR over time.
Cluster analysis was used to group patients with similar morphologic characteristics. Glomerular descriptors and patient clusters were assessed for associations with outcomes using adjusted Cox models and linear mixed models. Messenger RNA from glomerular tissue was used to assess differentially expressed genes between clusters and identify genes associated with individual descriptors driving cluster membership.
Three clusters were identified: X (n = 56), Y (n = 68), and Z (n = 97). Clusters Y and Z had higher probabilities of proteinuria remission (HRs of 1.95 95% CI, 0.99-3.85 and 3.29 95% CI, 1.52-7.13, respectively), lower hazards of disease progression (HRs of 0.22 95% CI, 0.08-0.57 and 0.11 95% CI, 0.03-0.45, respectively), and lower loss of eGFR over time compared with X. Cluster X had 1,920 genes that were differentially expressed compared with Y+Z; these reflected activation of pathways of immune response and inflammation. Six descriptors driving the clusters individually correlated with clinical outcomes and gene expression.
Low prevalence of some descriptors and biopsy at a single time point.
The NDPSS allows for categorization of FSGS/MCD patients into clinically and biologically relevant subgroups, and uncovers histologic parameters associated with clinical outcomes and molecular signatures not included in current classification systems.
Display omitted
Apolipoprotein L1 (APOL1)-associated focal segmental glomerulosclerosis (FSGS) is the dominant form of FSGS in Black individuals. There are no targeted therapies for this condition, in part because ...the molecular mechanisms underlying APOL1’s pathogenic contribution to FSGS are incompletely understood. Studying the transcriptomic landscape of APOL1 FSGS in patient kidneys is an important way to discover genes and molecular behaviors that are unique or most relevant to the human disease. With the hypothesis that the pathology driven by the high-risk APOL1 genotype is reflected in alteration of gene expression across the glomerular transcriptome, we compared expression and co-expression profiles of 15,703 genes in 16 Black patients with FSGS at high-risk vs 14 Black patients with a low-risk APOL1 genotype. Expression data from APOL1-inducible HEK293 cells and normal human glomeruli were used to pursue genes and molecular pathways uncovered in these studies. We discovered increased expression of APOL1 and nine other significant differentially expressed genes in high-risk patients. This included stanniocalcin, which has a role in mitochondrial and calcium-related processes along with differential correlations between high- and low-risk APOL1 and metabolism pathway genes. There were similar correlations with extracellular matrix- and immune-related genes, but significant loss of co-expression of mitochondrial genes in high-risk FSGS, and an NF-κB-down regulating gene, NKIRAS1, as the most significant hub gene with strong differential correlations with NDUF family (mitochondrial respiratory genes) and immune-related (JAK-STAT) genes. Thus, differences in mitochondrial gene regulation appear to underlie many differences observed between high- and low-risk Black patients with FSGS.
Display omitted
Background
In the current study, longitudinal BP and lipid measurements were examined in a NEPTUNE cohort of children with newly diagnosed nephrotic syndrome (cNEPTUNE). We hypothesized that ...hypertensive BP and dyslipidemia would persist in children with nephrotic syndrome, regardless of steroid treatment response.
Methods
A multi-center longitudinal observational analysis of data obtained from children < 19 years of age with new onset nephrotic syndrome enrolled in the Nephrotic Syndrome Study Network (cNEPTUNE) was conducted. BP and lipid data were examined over time stratified by disease activity and steroid exposure. Generalized estimating equation regressions were used to find determinants of hypertensive BP and dyslipidemia.
Results
Among 122 children, the prevalence of hypertensive BP at any visit ranged from 17.4% to 57.4%, while dyslipidemia prevalence ranged from 40.0% to 96.2% over a median of 30 months of follow-up. Hypertensive BP was found in 46.2% (116/251) of study visits during active disease compared with 31.0% (84/271) of visits while in remission. Dyslipidemia was present in 88.2% (120/136) of study visits during active disease and in 66.0% (101/153) while in remission. Neither dyslipidemia nor hypertensive BP were significantly different with/without medication exposure (steroids and/or CNI). In regression analysis, male sex and urine protein:creatinine ratio (UPC) were significant determinants of hypertensive BP over time, while eGFR was found to be a determinant of dyslipidemia over time.
Conclusions
Results demonstrate persistent hypertensive BPs and unfavorable lipid profiles in the cNEPTUNE cohort regardless of remission status or concurrent steroid or calcineurin inhibitor treatment.
Graphical abstract
A higher resolution version of the Graphical abstract is available as
Supplementary information
Steroids, the mainstay of treatment for nephrotic syndrome in children, have multiple adverse effects including growth suppression.
Anthropometric measurements in children < 18 years enrolled in the ...Nephrotic Syndrome Study Network (NEPTUNE) were collected. The longitudinal association of medication exposure and nephrotic syndrome characteristics with height z-score and growth velocity was determined using adjusted Generalized Estimating Equation regression and linear regression.
A total of 318 children (57.2% males) with a baseline age of 7.64 ± 5.04 years were analyzed. The cumulative steroid dose was 216.4 (IQR 61.5, 652.7) mg/kg (N = 233). Overall, height z-scores were not significantly different at the last follow-up compared to baseline (- 0.13 ± 1.21 vs. - 0.23 ± 1.71, p = 0.21). In models adjusted for age, sex, and eGFR, greater cumulative steroid exposure (β - 7.5 × 10
, CI - 1.2 × 10
, - 3 × 10
, p = 0.001) and incident cases of NS (vs. prevalent) (β - 1.1, CI - 2.22, - 0.11, p = 0.03) were significantly associated with lower height z-scores over time. Rituximab exposure was associated with higher height z-scores (β 0.16, CI 0.04, 0.29, p = 0.01) over time.
Steroid dose was associated with lower height z-score, while rituximab use was associated with higher height z-score.
Podocyte injury is central to many forms of kidney disease, but transcriptional signatures reflecting podocyte injury and compensation mechanisms are challenging to analyze in vivo. Human kidney ...organoids derived from pluripotent stem cells (PSCs), a potentially new model for disease and regeneration, present an opportunity to explore the transcriptional plasticity of podocytes. Here, transcriptional profiling of more than 12,000 single cells from human PSC-derived kidney organoid cultures was used to identify robust and reproducible cell lineage gene expression signatures shared with developing human kidneys based on trajectory analysis. Surprisingly, the gene expression signature characteristic of developing glomerular epithelial cells was also observed in glomerular tissue from a kidney disease cohort. This signature correlated with proteinuria and inverse eGFR, and it was confirmed in an independent podocytopathy cohort. Three genes in particular were further characterized as potentially novel components of the glomerular disease signature. We conclude that cells in human PSC-derived kidney organoids reliably recapitulate the developmental transcriptional program of podocytes and other cell lineages in the human kidney and that transcriptional profiles seen in developing podocytes are reactivated in glomerular disease. Our findings demonstrate an approach to identifying potentially novel molecular programs involved in the pathogenesis of glomerulopathies.
Abstract
Background
Understanding the relationship between clinical and patient-reported outcomes (PROs) will help support clinical care and future clinical trial design of novel therapies for focal ...segmental glomerulosclerosis (FSGS).
Methods
FSGS patients ≥8 years of age enrolled in the Nephrotic Syndrome Study Network completed Patient-Reported Outcomes Measurement Information System PRO measures of health-related quality of life (HRQoL) (children: global health, mobility, fatigue, pain interference, depression, anxiety, stress and peer relationships; adults: physical functioning, fatigue, pain interference, sleep impairment, mental health, depression, anxiety and social satisfaction) at baseline and during longitudinal follow-up for a maximum of 5 years. Linear mixed-effects models were used to determine which demographic, clinical and laboratory features were associated with PROs for each of the eight children and eight adults studied.
Results
There were 45 children and 114 adult FSGS patients enrolled that had at least one PRO assessment and 519 patient visits. Multivariable analyses among children found that edema was associated with global health (−7.6 points, P = 0.02) and mobility (−4.2, P = 0.02), the number of reported symptoms was associated with worse depression (−2.7 per symptom, P = 0.009) and anxiety (−2.3, P = 0.02) and the number of emergency room (ER) visits in the prior 6 months was associated with worse mobility (−2.8 per visit, P < 0.001) and fatigue (−2.4, P = 0.03). Multivariable analyses among adults found the number of reported symptoms was associated with worse function in all eight PROMIS measures and the number of ER visits was associated with worse fatigue, pain interference, sleep impairment, depression, anxiety and social satisfaction. Laboratory markers of disease severity (i.e. proteinuria, estimated glomerular filtration rate and serum albumin) did not predict PRO in multivariable analyses, with the single exception of complete remission and better pain interference scores among children (+9.3, P = 0.03).
Conclusions
PROs provide important information about HRQoL for persons with FSGS that is not captured solely by the examination of laboratory-based markers of disease. However, it is critical that instruments capture the patient experience and FSGS clinical trials may benefit from a disease-specific instrument more sensitive to within-patient changes.
Diabetic nephropathy (DN), a major cause of ESRD, is undoubtedly multifactorial and is caused by environmental and genetic factors. To identify a genetic basis for DN susceptibility, we are ...collecting multiplex DN families in the Caucasian (CA) and African-American (AA) populations for whole genome scanning and candidate gene analysis. A candidate gene search of diabetic sibs discordantly affected, concordantly affected and concordantly unaffected for DN was performed with microsatellite markers in genomic regions suspected to harbor nephropathy susceptibility loci. Regions examined were at human chromosome 10p,10q (orthologous to the rat renal susceptibility Rf-1 locus), and at NPHS1 (nephrin), CD2AP, Wilms tumor (WT1), and NPHS2 (podocin) loci. Linkage analyses were conducted using model-free methods (SIBPAL, S.A.G.E.) for AA, CA, and the combined sample. Allele frequencies and the identity by descent sharing were estimated separately for AA and CA, and race was included as a covariate in the final linkage analysis. To date, we have collected 212 sib pairs from 46 CA and 50 AA families. The average age of diabetes onset was 46.8 yr versus 36.2 yr for CA and 39.5 yr versus 40.2 yr for AA, in males versus females respectively. Genotyping data were available for 106 sib pairs (43 CA, 63 AA) from 27 CA (44% male probands) and 38 AA families (43% male probands). Average AA and CA sibship size was 2.73. Singlepoint and multipoint linkage analyses indicate that marker D10S1654 on chromosome 10p is potentially linked to DN (CA only multipoint P = 4 x 10(-3)). Interestingly, the majority of the linkage evidence derives from the CA sib pairs. We are now adding sib pairs and increasing marker density on chromosome 10. We have excluded linkage with candidate regions for nephrin, CD2AP, WT1, and podocin in this sample. In conjunction with previous reports, our data support evidence for a DN susceptibility locus on chromosome 10.
We conducted this study to test the hypothesis that plasma zonulin levels are elevated in pediatric patients with nephrotic syndrome compared to healthy controls.
Plasma zonulin levels were measured ...by ELISA in 114 children enrolled in the NEPTUNE study. Clinical and laboratory data were retrieved from the NEPTUNE database.
The median age of the patients was 10 (IQR = 5 to 14) years, 59 were male, 64 had minimal change disease, 47 focal segmental glomerulosclerosis, median eGFR was 96 (IQR = 80 to 114) ml/min/1.73 m
, and median urine protein:creatinine ratio was 0.5 (IQR = 0.1 to 3.4) (g:g). The plasma zonulin level was 14.2 ± 5.0 vs. 10.2 ± 2.5 ng/ml in healthy adults in a report using the same assay kit,
= 0.0025. These findings were confirmed in an independent cohort of children with nephrotic syndrome compared to healthy age-matched controls,
= 0.01. Zonulin concentrations did not differ in children with minimal change disease vs. focal segmental glomerulosclerosis, frequently relapsing vs. steroid-dependent vs. steroid-resistant clinical course, and were not influenced by the immunosuppressive treatment regimen. There was no relationship between plasma zonulin levels and the absolute or percentage change in proteinuria from enrollment until the time of the zonulin assay.
Plasma zonulin levels are elevated in childhood nephrotic syndrome regardless of level of proteinuria or specific treatment. The cause of the high plasma zonulin levels and whether zonulin contributes to glomerular injury requires further study.