Automatic Evaluation of Histological Prognostic Factors Using Two Consecutive Convolutional Neural Networks on Kidney Samples

Background and Objectives: The prognosis of patients undergoing kidney tumor resection or kidney donation is linked to many histological criteria. These criteria notably include glomerular density, glomerular volume, vascular luminal stenosis, and severity of interstitial fibrosis/tubular atrophy. Automated measurements through a Deep Learning approach could save time and provide more precise data. This work aimed to develop a free tool to automatically obtain kidney histological prognostic features. Design, setting, participants, and measurements: Two hundred and forty one samples of healthy kidney tissue were split into 3 independent cohorts. The "Training" cohort (n=65) was used to train two Convolutional Neural Networks: one to detect the cortex and a second one to segment the kidney structures. The "Test" cohort (n=50) assessed their performances by comparing manually outlined regions of interest to predicted ones. The "Application" cohort (n=126) compared prognostic histological data obtained manually or through the algorithm based on the combination of the two Convolutional Neural Networks. Results: In the Test cohort, the networks isolated the cortex and segmented the elements of interest with good performances (more than 90% of the cortex, healthy tubules, glomeruli, and even globally sclerotic glomeruli were detected). In the Application cohort, the expected and predicted prognostic data were significantly correlated. The correlation coefficients r were respectively 0.85 for glomerular volume, 0.51 for glomerular density, 0.75 for interstitial fibrosis, 0.71 for tubular atrophy, and 0.73 for vascular intimal thickness. The algorithm had a good ability to predict significant (> 25%) tubular atrophy and interstitial fibrosis level (ROC curve with an area under the curve 0.92 and 0.91, respectively) or a significant vascular luminal stenosis (> 50%) (area under the curve 0.85). Conclusion: This freely available tool enables the automated segmentation of kidney tissue to obtain prognostic histological data in a fast, objective, reliable and reproducible way.

Podocyte Foot Process Effacement Precedes Albuminuria and Glomerular Hypertrophy in CD2-Associated Protein Deficient Mice

Background: Podocyte foot process effacement is a key histologic finding in proteinuric kidney disease. We previously showed that 3-week old CD2AP-deficient mice have significant proteinuria, glomerular hypertrophy and mesangial expansion. The goal of this study is to use morphometry to establish the temporal sequence of podocyte foot process effacement, glomerular volume expansion and albuminuria in Cd2ap−/− mice by measuring these parameters at the 2-week time point.Methods: Wild-type mice age 14 ± 1 days with the Cd2ap gene (WT, N = 5) and mice deficient for Cd2ap (Cd2ap KO, N = 5) were generated. Kidneys were harvested and fixed in 2.5% glutaraldehyde and processed for examination by light and electron microscopy. An average of 415.2 (range 268–716) grid points were counted for all the glomeruli, and quantification of glomerular volume from each kidney. Urine was collected the day prior to sacrifice for urine albumin-to-creatinine ratio (ACR) measurements.Results: There was no difference in albuminuria [median (range) mg/g] between WT [212.2 (177.6–388.4) mg/g] vs. Cd2ap KO mice [203.3 (164.7–910.2) mg/g], P = 0.89; or glomerular volume 68,307[10,931] vs. 66,844[13,022] μm3, p = 0.92. The volume densities of glomerular components of the podocyte, capillary lumen and mesangium were not different for the two groups, P = 0.14, 0.14 and 0.17 respectively. However, foot process width was increased in Cd2ap KO 1128[286] vs. WT [374 ± 42] nm, P = 0.02.Conclusion: Here we show that while 2-week old WT and Cd2ap KO mice have similar levels of albuminuria, glomerular and mesangial volume, Cd2ap KO mice have more extensive podocyte foot process effacement. The data suggests that podocyte injury is the initiating event leading to mesangial expansion and albuminuria in this model.

AMP-Kinase mediates regulation of glomerular volume and podocyte survival

We reported that Shroom3 knockdown, via Fyn inhibition, induced albuminuria with foot process effacement (FPE) without glomerulosclerosis (FSGS) or podocytopenia. Interestingly, knockdown mice had reduced podocyte volumes. Human minimal change disease, where podocyte Fyn inactivation was reported, also showed lower glomerular volumes than FSGS. We hypothesized that lower glomerular volume prevented the progression to podocytopenia. To test this hypothesis, we utilized unilateral- and 5/6th nephrectomy models in Shroom3 knockdown mice. Knockdown mice exhibited lower glomerular volume, and less glomerular and podocyte hypertrophy after nephrectomy. FYN-knockdown podocytes had similar reductions in podocyte volume, implying Fyn was downstream of Shroom3. Using SHROOM3- or FYN-knockdown, we confirmed reduced podocyte protein content, along with significantly increased phosphorylated AMP-kinase, a negative regulator of anabolism. AMP-Kinase activation resulted from increased cytoplasmic redistribution of LKB1 in podocytes. Inhibition of AMP-Kinase abolished the reduction in glomerular volume and induced podocytopenia in mice with FPE, suggesting a protective role for AMP-Kinase activation. In agreement with this, treatment of glomerular injury models with AMP-Kinase activators restricted glomerular volume, podocytopenia and progression to FSGS. In summary, we demonstrate the important role of AMP-Kinase in glomerular volume regulation and podocyte survival. Our data suggest that AMP-Kinase activation adaptively regulates glomerular volume to prevent podocytopenia in the context of podocyte injury.

MO089HISTOMORPHOMETRIC ANALYSIS OF GLOMERULI AND CAPILLARIES DENSITY IN THE INTERSTITIUM IN PREIMPLANTATION NEEDLE BIOPSIES OF PAIRED KIDNEYS HARVESTED FROM ADULT CADAVERIC DONORS

Background and Aims Preimplantation needle biopsy of kidney allows more precise interpretation of subsequent kidney biopsies performed after transplantation and potentially also may predict kidney graft survival. It is unknown, whether it is justified to perform the biopsy of one kidney only and then to transmit obtained results also to the second kidney or biopsies of both kidneys are mandatory. The aim of the study was to assess differences regarding glomerular volume, glomerular density and capillary density in the interstitium of both kidneys harvested from the same deceased donor. Method The study involved 40 pairs of kidneys (all together 80 kidneys) in which preimplantation kidney biopsies were performed. Kidneys were harvested from 40 deceased donors (17 females and 23 males; mean age 42.3 [37.6-47.0] years old) died because of intracranial haemorrhage (17 donors) or cerebral trauma due to accident (23 donors). Histomorphometric analysis was performed using “Olympus BX51” microscope (Olympus, Tokyo, Japan) coupled with “Olympus BX50” camera and the “cellSens Standard” software (Olympus, Tokyo, Japan). Weibel-Gomez formula was adapted to calculate glomerular volume. Results No significant differences were found between mean kidneys length [115.8 (112.4-119.1) vs.115.5 (112.5-118.5) mm], glomerular volume [2.59 (2.24-2.93) vs 2.49 (2.15-2.84) μm3 x106 ], glomerular density [3.43 (3.07-3.80) vs 3.24 (2.87-3.61) n/mm2] and interstitial capillaries density [233.58 (211.26) vs 217.80 (199.45-236.47) n/mm2] of both kidneys harvested from the same deceased donor. Conclusions 1. Both kidneys harvested from the same deceased donor did not differ significantly in kidney length, glomerular volume, glomerular density and capillary density. 2. Our results justify to preimplantation biopsy of only one kidney and the results from the histomorphometric analysis may be used in the future also for assessment of the second kidney.

Larger Nephron Size and Nephrosclerosis Predict Progressive CKD and Mortality after Radical Nephrectomy for Tumor and Independent of Kidney Function

BackgroundNephron hypertrophy and nephrosclerosis may be important determinants of CKD and mortality. However, studies of outcomes associated with these microstructural features have been limited to small tissue specimens from patients selected for either good kidney health or known kidney disease.MethodsTo determine whether microstructural features are predictive of progressive CKD and mortality outcomes, we studied patients who underwent a radical nephrectomy for a tumor. Large wedge sections of renal parenchyma distal to the tumor were stained and scanned into high-resolution images; we annotated the cortex and all glomeruli to calculate glomerular volume, cortex volume per glomerulus, and percentage of globally sclerotic glomeruli. Morphometric measurements also included percentages of artery luminal stenosis and interstitial fibrosis/tubular atrophy (IF/TA) of the cortex. At follow-up visits every 6–12 months, we determined which patients experienced progressive CKD (defined as dialysis, kidney transplantation, or a 40% decline from postnephrectomy eGFR). Cox models for these outcomes were adjusted for age, sex, body mass index, hypertension, diabetes, smoking, eGFR, and proteinuria.ResultsAmong 936 patients (mean age, 64 years; postnephrectomy baseline eGFR, 48 ml/min per 1.73 m2), 117 progressive CKD events, 183 noncancer deaths, and 116 cancer deaths occurred during a median follow-up of 6.4 years. Larger glomerular volume, larger cortex per glomerulus, and higher percentage of globally sclerotic glomeruli or IF/TA predicted progressive CKD. Higher percentage IF/TA also predicted noncancer mortality. Microstructural features did not predict cancer mortality or recurrence.ConclusionsAfter a radical nephrectomy, larger nephrons and nephrosclerosis predicted progressive CKD, and IF/TA predicted noncancer mortality. Morphometric analysis of renal parenchyma can predict noncancer clinical events in patients long after their radical nephrectomy.

Automated Image Analyses of Glomerular Hypertrophy in a Mouse Model of Diabetic Nephropathy

BackgroundGlomerular hypertrophy is a hallmark of kidney injury in metabolically induced renal diseases such as obesity-associated glomerulopathies and diabetic nephropathy (DN).MethodsUsing light sheet fluorescent microscopy (LSFM) and 3D image analysis, we tested algorithms for automated and unbiased quantification of total glomerular numbers and individual glomerular volume in the uninephrectomized (UNx) db/db mouse model of DN.ResultsAt 6 weeks after surgery, db/db and UNx db/db mice showed increased urine albumin-to-creatinine ratio (ACR) compared with db/+ control mice. Before euthanasia, glomeruli were labeled in vivo by injecting tomato lectin. Whole-kidney LSFM 3D image analysis revealed that mean glomerular volume was significantly increased in UNx db/db mice compared with db/+ mice. Moreover, analysis of individual glomerular volume showed a shift in volume distribution toward larger glomeruli and thereby demonstrated additive effects of diabetes and UNx on induction of glomerular hypertrophy. The automatized quantification showed no significant differences in glomerular numbers among db/+, db/db, and UNx db/db mice. These data correlated with glomerular numbers as quantified by subsequent stereologic quantification.ConclusionsOverall, LSFM coupled with automated 3D histomorphometric analysis was demonstrated to be advantageous for unbiased assessment of glomerular volume and numbers in mouse whole-kidney samples. Furthermore, we showed that injection of fluorescently labeled lectin and albumin can be used as markers of nephron segments in the mouse kidneys, thus enabling functional assessment of kidney physiology, pathology, and pharmacology in preclinical rodent models of kidney disease.

Podocyte hypertrophic stress and detachment precedes hyperglycemia or albuminuria in a rat model of obesity and type2 diabetes-associated nephropathy

Type2 diabetes-associated nephropathy is the commonest cause of renal failure. Mechanisms responsible are controversial. Leptin-deficient hyperphagic Zucker (fa/fa) rats were modeled to test the hypothesis that glomerular enlargement drives podocyte hypertrophic stress leading to accelerated podocyte detachment, podocyte depletion, albuminuria and progression. By 6weeks, prior to development of either hyperglycemia or albuminuria, fa/fa rats were hyperinsulinemic with high urinary IGF1/2 excretion, gaining weight rapidly, and had 1.6-fold greater glomerular volume than controls (P < 0.01). At this time the podocyte number per glomerulus was not yet reduced although podocytes were already hypertrophically stressed as shown by high podocyte phosphor-ribosomal S6 (a marker of mTORC1 activation), high urinary pellet podocin:nephrin mRNA ratio and accelerated podocyte detachment (high urinary pellet podocin:aquaporin2 mRNA ratio). Subsequently, fa/fa rats became both hyperglycemic and albuminuric. 24 hr urine albumin excretion correlated highly with decreasing podocyte density (R2 = 0.86), as a consequence of both increasing glomerular volume (R2 = 0.70) and decreasing podocyte number (R2 = 0.63). Glomerular podocyte loss rate was quantitatively related to podocyte detachment rate measured by urine pellet mRNAs. Glomerulosclerosis occurred when podocyte density reached <50/106um3. Reducing food intake by 40% to slow growth reduced podocyte hypertrophic stress and “froze” all elements of the progression process in place, but had small effect on hyperglycemia. Glomerular enlargement caused by high growth factor milieu starting in pre-diabetic kidneys appears to be a primary driver of albuminuria in fa/fa rats and thereby an under-recognized target for progression prevention. Progression risk could be identified prior to onset of hyperglycemia or albuminuria, and monitored non-invasively by urinary pellet podocyte mRNA markers.

Glomerular Volume and Glomerulosclerosis at Different Depths within the Human Kidney

BackgroundAge, CKD risk factors, and kidney function are associated with larger glomerular volume and a higher percentage of globally sclerotic glomeruli. Knowledge of how these associations may differ by cortical depth is limited.MethodsTo investigate glomerular volume and glomerulosclerosis across different depths of cortex, we studied wedge sections of the renal parenchyma from 812 patients who underwent a radical nephrectomy (for a tumor), separately characterizing glomeruli in the superficial (subcapsular), middle, and deep (juxtamedullary) regions. We compared the association of mean nonsclerotic glomerular volume and of glomerulosclerosis (measured as the percentage of globally sclerotic glomeruli) with age, obesity, diabetes, smoking, kidney function, and structural pathology in the superficial, middle, and deep regions.ResultsThe superficial, middle, and deep regions showed significant differences in glomerular volume (0.0025, 0.0031, and 0.0028 µm3, respectively) and in glomerulosclerosis (18%, 7%, and 11%, respectively). There was a marked increase in glomerulosclerosis with age in the superficial region, but larger glomerular volume was not associated with age at any cortical depth. Glomerulosclerosis associated more strongly with arteriosclerosis and ischemic-appearing glomeruli in the superficial region. Hypertension, lower eGFR, and interstitial fibrosis associated with glomerulosclerosis and glomerular volume to a similar extent at any depth. Diabetes and proteinuria more strongly associated with glomerulosclerosis in the deep and middle regions, respectively, but neither associated with glomerular volume differently by depth. Obesity associated more strongly with glomerular volume in the superficial cortex.ConclusionsMost clinical characteristic show similar associations with glomerulosclerosis and glomerulomegaly at different cortical depths. Exceptions include age-related glomerulosclerosis, which appears to be an ischemic process and is more predominant in the superficial region.