Performance of an Artificial Intelligence Model for Recognition and Quantitation of Histologic Features of Eosinophilic Esophagitis on Biopsy Samples.

We have developed an artificial intelligence (AI)-based digital pathology model for the evaluation of histologic features related to eosinophilic esophagitis (EoE). In this study, we evaluated the performance of our AI model in a cohort of pediatric and adult patients for histologic features included in the Eosinophilic Esophagitis Histologic Scoring System (EoEHSS). We collected a total of 203 esophageal biopsy samples from patients with mucosal eosinophilia of any degree (91 adult and 112 pediatric patients) and 10 normal controls from a prospectively maintained database. All cases were assessed by a specialized gastrointestinal (GI) pathologist for features in the EoEHSS at the time of original diagnosis and rescored by a central GI pathologist (R.K.M.). We subsequently analyzed whole-slide image digital slides using a supervised AI model operating in a cloud-based, deep learning AI platform (Aiforia Technologies) for peak eosinophil count (PEC) and several histopathologic features in the EoEHSS. The correlation and interobserver agreement between the AI model and pathologists (Pearson correlation coefficient [rs] = 0.89 and intraclass correlation coefficient [ICC] = 0.87 vs original pathologist; rs = 0.91 and ICC = 0.83 vs central pathologist) were similar to the correlation and interobserver agreement between pathologists for PEC (rs = 0.88 and ICC = 0.91) and broadly similar to those for most other histologic features in the EoEHSS. The AI model also accurately identified PEC of >15 eosinophils/high-power field by the original pathologist (area under the curve [AUC] = 0.98) and central pathologist (AUC = 0.98) and had similar AUCs for the presence of EoE-related endoscopic features to pathologists' assessment. Average eosinophils per epithelial unit area had similar performance compared to AI high-power field-based analysis. Our newly developed AI model can accurately identify, quantify, and score several of the main histopathologic features in the EoE spectrum, with agreement regarding EoEHSS scoring which was similar to that seen among GI pathologists.

A Higher Foci Density of Interstitial Fibrosis and Tubular Atrophy Predicts Progressive CKD after a Radical Nephrectomy for Tumor.

BACKGROUND: Chronic tubulointerstitial injury on kidney biopsy is usually quantified by the percentage of cortex with interstitial fibrosis/tubular atrophy (IF/TA). Whether other patterns of IF/TA or inflammation in the tubulointerstitium have prognostic importance beyond percentage IF/TA is unclear. METHODS: We obtained, stained, and digitally scanned full cortical thickness wedge sections of renal parenchyma from patients who underwent a radical nephrectomy for a tumor over 2000-2015, and morphometrically analyzed the tubulointerstitium of the cortex for percentage IF/TA, IF/TA density (foci per mm2 cortex), percentage subcapsular IF/TA, striped IF/TA, percentage inflammation (both within and outside IF/TA regions), and percentage subcapsular inflammation. Patients were followed with visits every 6-12 months. Progressive CKD was defined as dialysis, kidney transplantation, or 40% decline from the postnephrectomy eGFR. Cox models assessed the risk of CKD or noncancer mortality with morphometric measures of tubulointerstitial injury after adjustment for the percentage IF/TA and clinical characteristics. RESULTS: Among 936 patients (mean age, 64 years; postnephrectomy baseline eGFR, 48 ml/min per 1.73m2), 117 progressive CKD events and 183 noncancer deaths occurred over a median 6.4 years. Higher IF/TA density predicted both progressive CKD and noncancer mortality after adjustment for percentage IF/TA and predicted progressive CKD after further adjustment for clinical characteristics. Independent of percentage IF/TA, age, and sex, higher IF/TA density correlated with lower eGFR, smaller nonsclerosed glomeruli, more global glomerulosclerosis, and smaller total cortical volume. CONCLUSIONS: Higher density of IF/TA foci (a more scattered pattern with more and smaller foci) predicts higher risk of progressive CKD after radical nephrectomy compared with the same percentage of IF/TA but with fewer and larger foci.

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

BACKGROUND: Nephron 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. METHODS: To 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. RESULTS: Among 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. CONCLUSIONS: After 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.