PICaSSO Histologic Remission Index (PHRI) in ulcerative colitis: development of a novel simplified histological score for monitoring mucosal healing and predicting clinical outcomes and its applicability in an artificial intelligence system.

Histological remission is evolving as an important treatment target in UC. We aimed to develop a simple histological index, aligned to endoscopy, correlated with clinical outcomes, and suited to apply to an artificial intelligence (AI) system to evaluate inflammatory activity. METHODS: Using a set of 614 biopsies from 307 patients with UC enrolled into a prospective multicentre study, we developed the Paddington International virtual ChromoendoScopy ScOre (PICaSSO) Histologic Remission Index (PHRI). Agreement with multiple other histological indices and validation for inter-reader reproducibility were assessed. Finally, to implement PHRI into a computer-aided diagnosis system, we trained and tested a novel deep learning strategy based on a CNN architecture to detect neutrophils, calculate PHRI and identify active from quiescent UC using a subset of 138 biopsies. RESULTS: PHRI is strongly correlated with endoscopic scores (Mayo Endoscopic Score and UC Endoscopic Index of Severity and PICaSSO) and with clinical outcomes (hospitalisation, colectomy and initiation or changes in medical therapy due to UC flare-up). A PHRI score of 1 could accurately stratify patients' risk of adverse outcomes (hospitalisation, colectomy and treatment optimisation due to flare-up) within 12 months. Our inter-reader agreement was high (intraclass correlation 0.84). Our preliminary AI algorithm differentiated active from quiescent UC with 78% sensitivity, 91.7% specificity and 86% accuracy. CONCLUSIONS: PHRI is a simple histological index in UC, and it exhibits the highest correlation with endoscopic activity and clinical outcomes. A PHRI-based AI system was accurate in predicting histological remission.

Expression of neutrophil gelatinase-associated lipocalin (NGAL) in the gut in Crohn's disease.

The antimicrobial glycoprotein neutrophil gelatinase-associated lipocalin (NGAL) is strongly expressed in several infectious, inflammatory and malignant disorders, among these inflammatory bowel disease (IBD). Fecal and serum NGAL is elevated during active IBD and we have recently shown that fecal NGAL is a novel biomarker for IBD with a test performance comparable to the established fecal biomarker calprotectin. This study examines expression of NGAL in the healthy gut and in Crohn's disease (CD), with emphasis on the previously unexplored small intestine. Pinch biopsies were taken from active and inactive CD in jejunum, ileum and colon and from the same sites in healthy controls. Microarray gene expression showed that the NGAL gene, LCN2, was the second most upregulated among 1820 differentially expressed genes in terminal ileum comparing active CD and controls (FC 5.86, p = 0.027). Based on immunohistochemistry and in situ hybridization findings, this upregulation most likely represented increased expression in epithelial cells. Double immunofluorescence showed NGAL expression in 49% (range 19-70) of Paneth cells (PCs) in control ileum with no change during inflammation. In healthy jejunum, the NGAL expression in PCs was weak to none but markedly increased during active CD. We further found NGAL also in metaplastic PCs in colon. Finally, we show for the first time that NGAL is expressed in enteroendocrine cells in small intestine as well as in colon.

Deep learning-based image analysis reveals significant differences in the number and distribution of mucosal CD3 and γδ T cells between Crohn's disease and ulcerative colitis.

Colon mucosae of ulcerative colitis (UC) and Crohn's disease (CD) display differences in the number and distribution of immune cells that are difficult to assess by eye. Deep learning-based analysis on whole slide images (WSIs) allows extraction of complex quantitative data that can be used to uncover different inflammatory patterns. We aimed to explore the distribution of CD3 and γδ T cells in colon mucosal compartments in histologically inactive and active inflammatory bowel disease. By deep learning-based segmentation and cell detection on WSIs from a well-defined cohort of CD (n = 37), UC (n = 58), and healthy controls (HCs, n = 33), we quantified CD3 and γδ T cells within and beneath the epithelium and in lamina propria in proximal and distal colon mucosa, defined by the Nancy histological index. We found that inactive CD had significantly fewer intraepithelial γδ T cells than inactive UC, but higher total number of CD3 cells in all compartments than UC and HCs. Disease activity was associated with a massive loss of intraepithelial γδ T cells in UC, but not in CD. The total intraepithelial number of CD3 cells remained constant regardless of disease activity in both CD and UC. There were more mucosal CD3 and γδ T cells in proximal versus distal colon. Oral corticosteroids had an impact on γδ T cell numbers, while age, gender, and disease duration did not. Relative abundance of γδ T cells in mucosa and blood did not correlate. This study reveals significant differences in the total number of CD3 and γδ T cells in particularly the epithelial area between CD, UC, and HCs, and demonstrates useful application of deep segmentation to quantify cells in mucosal compartments.

Response Evaluation after Neoadjuvant Chemotherapy for Resectable Gastric Cancer.

BACKGROUND: The method of response evaluation following neoadjuvant chemotherapy (NAC) in resectable gastric cancer has been widely debated. An essential prerequisite is the ability to stratify patients into subsets of different long-term survival rates based on the response mode. Histopathological measures of regression have their limitations, and interest resides in CT-based methods that can be used in everyday settings. METHODS: We conducted a population-based study (2007-2016) on 171 consecutive patients with gastric adenocarcinoma who were receiving NAC. Two methods of response evaluation were investigated: a strict radiological procedure using RECIST (downsizing), and a composite radiological/pathological procedure comparing the initial radiological TNM stage to the pathological ypTNM stage (downstaging). Clinicopathological variables that could predict the response were searched for, and correlations between the response mode and long-term survival rates were assessed. RESULTS: RECIST failed to identify half of the patients progressing to metastatic disease, and it was unable to assign patients to subsets with different long-term survival rates based on the response mode. However, the TNM stage response mode did achieve this objective. Following re-staging, 48% (78/164) were downstaged, 15% (25/164) had an unchanged stage, and 37% (61/164) were upstaged. A total of 9% (15/164) showed a histopathological complete response. The 5-year overall survival rate was 65.3% (95% CI 54.7-75.9%) for TNM downstaged cases, 40.0% (95% CI 20.8-59.2%) for stable disease, and 14.8% (95% CI 6.0-23.6%) for patients with TNM progression, p < 0.001. In a multivariable ordinal regression model, the Lauren classification and tumor site were the only significant determinants of the response mode. CONCLUSIONS: Downsizing, as a method for evaluating the response to NAC in gastric cancer, is discouraged. TNM re-staging by comparing the baseline radiological CT stage to the pathological stage following NAC is suggested as a useful method that may be used in everyday situations.

Profound gene expression changes in the epithelial monolayer of active ulcerative colitis and Crohn's disease.

In recent years it has become apparent that the epithelium is highly involved in inflammatory bowel disease (IBD) pathophysiology. The majority of gene expression studies of IBD are generated from heterogeneous biopsies, providing no distinction between immune cells, the epithelium and other mucosal cells. By using laser capture microdissection (LCM) coupled with RNA sequencing, we aimed to characterize the expressional changes of the isolated colonic epithelial monolayer from ulcerative colitis (UC) and Crohn's disease (CD) patients compared to healthy controls (HC). The analysis identified 3706 genes as differentially expressed between active IBD epithelium and HC. Weighted gene co-expression network analysis was used to stratify genes into modules, which were subsequently characterized using enrichment analysis. Our data show a distinct upregulation of the antigen presentation machinery during inflammation, including major histocompatibility complex class II molecules (e.g. HLA-DPA1, HLA-DPB1, HLA-DRA) and key transcription factors/activators (STAT1, IRF1, CIITA). We also see an epithelial downregulation of retinoic acid-responsive nuclear receptors (RARA, RARB, RXRA), but upregulation of retinoid-metabolizing enzymes (RDH11, ALDH1A2, ALDH1A3), which together suggest a perturbation of epithelial vitamin A signaling during active IBD. Lastly, we identified a cluster of stress-related genes, including activator protein 1 components JUNB and ATF3, as significantly upregulated in active UC but not in CD, revealing an interesting aspect of IBD heterogeneity. The results represent a unique resource for enhanced understanding of epithelial involvement in IBD inflammation and is a valuable tool for further studies on these processes.

Constrained multiple instance learning for ulcerative colitis prediction using histological images.

BACKGROUND AND OBJECTIVE: Ulcerative colitis (UC) is an inflammatory bowel disease (IBD) affecting the colon and the rectum characterized by a remitting-relapsing course. To detect mucosal inflammation associated with UC, histology is considered the most stringent criteria. In turn, histologic remission (HR) correlates with improved clinical outcomes and has been recently recognized as a desirable treatment target. The leading biomarker for assessing histologic remission is the presence or absence of neutrophils. Therefore, the finding of this cell in specific colon structures indicates that the patient has UC activity. However, no previous studies based on deep learning have been developed to identify UC based on neutrophils detection using whole-slide images (WSI). METHODS: The methodological core of this work is a novel multiple instance learning (MIL) framework with location constraints able to determine the presence of UC activity using WSI. In particular, we put forward an effective way to introduce constraints about positive instances to effectively explore additional weakly supervised information that is easy to obtain and enjoy a significant boost to the learning process. In addition, we propose a new weighted embedding to enlarge the relevance of the positive instances. RESULTS: Extensive experiments on a multi-center dataset of colon and rectum WSIs, PICASSO-MIL, demonstrate that using the location information we can improve considerably the results at WSI-level. In comparison with prior MIL settings, our method allows for 10% improvements in bag-level accuracy. CONCLUSION: Our model, which introduces a new form of constraints, surpass the results achieved from current state-of-the-art methods that focus on the MIL paradigm. Our method can be applied to other histological concerns where the morphological features determining a positive WSI are tiny and similar to others in the image.

Bacterial Mucosa-associated Microbiome in Inflamed and Proximal Noninflamed Ileum of Patients With Crohn's Disease.

BACKGROUND: Microbiota is most likely essential in the pathogenesis of Crohn's disease (CD). Fecal diversion after ileocecal resection (ICR) protects against CD recurrence, whereas infusion of fecal content triggers inflammation. After ICR, the majority of patients experience endoscopic recurrence in the neoterminal ileum, and the ileal microbiome is of particular interest. We have assessed the mucosa-associated microbiome in the inflamed and noninflamed ileum in patients with CD. METHODS: Mucosa-associated microbiome was assessed by 16S rRNA sequencing of biopsies sampled 5 and 15 cm orally of the ileocecal valve or ileocolic anastomosis. RESULTS: Fifty-one CD patients and forty healthy controls (HCs) were included in the study. Twenty CD patients had terminal ileitis, with endoscopic inflammation at 5 cm, normal mucosa at 15 cm, and no history of upper CD involvement. Crohn's disease patients (n = 51) had lower alpha diversity and separated clearly from HC on beta diversity plots. Twenty-three bacterial taxa were differentially represented in CD patients vs HC; among these, Tyzzerella 4 was profoundly overrepresented in CD. The microbiome in the inflamed and proximal noninflamed ileal mucosa did not differ according to alpha diversity or beta diversity. Additionally, no bacterial taxa were differentially represented. CONCLUSIONS: The microbiome is similar in the inflamed and proximal noninflamed ileal mucosa within the same patients. Our results support the concept of CD-specific microbiota alterations and demonstrate that neither ileal sublocation nor endoscopic inflammation influence the mucosa-associated microbiome.

Code-Free Development and Deployment of Deep Segmentation Models for Digital Pathology.

Application of deep learning on histopathological whole slide images (WSIs) holds promise of improving diagnostic efficiency and reproducibility but is largely dependent on the ability to write computer code or purchase commercial solutions. We present a code-free pipeline utilizing free-to-use, open-source software (QuPath, DeepMIB, and FastPathology) for creating and deploying deep learning-based segmentation models for computational pathology. We demonstrate the pipeline on a use case of separating epithelium from stroma in colonic mucosa. A dataset of 251 annotated WSIs, comprising 140 hematoxylin-eosin (HE)-stained and 111 CD3 immunostained colon biopsy WSIs, were developed through active learning using the pipeline. On a hold-out test set of 36 HE and 21 CD3-stained WSIs a mean intersection over union score of 95.5 and 95.3% was achieved on epithelium segmentation. We demonstrate pathologist-level segmentation accuracy and clinical acceptable runtime performance and show that pathologists without programming experience can create near state-of-the-art segmentation solutions for histopathological WSIs using only free-to-use software. The study further demonstrates the strength of open-source solutions in its ability to create generalizable, open pipelines, of which trained models and predictions can seamlessly be exported in open formats and thereby used in external solutions. All scripts, trained models, a video tutorial, and the full dataset of 251 WSIs with ~31 k epithelium annotations are made openly available at https://github.com/andreped/NoCodeSeg to accelerate research in the field.

Mucosal and faecal neutrophil gelatinase-associated lipocalin as potential biomarkers for collagenous colitis.

BACKGROUND: Collagenous colitis (CC) is an inflammatory bowel disease where chronic diarrhoea is the main symptom. Diagnostic markers distinguishing between CC and other causes of chronic diarrhoea remain elusive. This study explores neutrophil gelatinase-associated lipocalin (NGAL) and its mRNA lipocalin2 (LCN2) as histological and faecal disease markers in CC. METHODS: NGAL/LCN2 were studied in colonic biopsies from CC patients before and during budesonide treatment using RNA sequencing (n = 9/group), in situ hybridization (ISH) (n = 13-22/group) and immunohistochemistry (IHC) (n = 14-25/group). Faecal samples from CC (n = 3-28/group), irritable bowel syndrome diarrhoea (IBS-D) (n = 14) and healthy controls (HC) (n = 15) were assayed for NGAL and calprotectin. RESULTS: NGAL/LCN2 protein and mRNA expression were upregulated in active CC vs HC, and vs paired samples of treated CC in clinical remission. IHC and ISH localized increased NGAL/LCN2 mainly to epithelium of active CC, compared to almost absence in HC and treated CC. In contrast, calprotectin was solely expressed in immune cells. Despite great individual differences, faecal NGAL was significantly increased in active CC compared to HC, IBS-D and treated CC and had high test sensitivity. Faecal calprotectin levels were variably increased in active CC, but the values remained below usual clinical cut-offs. CONCLUSION: NGAL/LCN2 is upregulated in the epithelium of active CC and reduced during budesonide-induced clinical remission to the level of HC and IBD-S. This was reflected in NGAL faecal concentrations. We propose NGAL as an IHC marker for disease activity in CC and a potential faecal biomarker discriminating CC from HC and IBS-D.

Mucosal 5-aminosalicylic acid concentration, drug formulation and mucosal microbiome in patients with quiescent ulcerative colitis.

BACKGROUND: 5-aminosalicylic acid (5-ASA) is the first-line therapy for ulcerative colitis (UC). 5-ASA acts locally in the colonic mucosa by numerous proposed mechanisms, and is metabolised by N-acetyltransferase (NAT). Large variations in mucosal 5-ASA concentrations have been reported, but the underlying mechanisms are not understood. AIM: To study the relationship between 5-ASA concentration, 5-ASA formulation, NAT genotype and bacterial microbiome in patients with UC. METHODS: Patients with quiescent UC, using monotherapy of Mezavant (n = 18), Asacol (n = 14) or Pentasa (n = 10), 4.0-4.8 g/day were included. 5-ASA was measured in colonic mucosal biopsies and serum by ultra-high performance liquid chromatography. NAT genotypes were determined by Sanger sequencing. Bacterial microbiome was sequenced from faeces and mucosa by 16S rRNA sequencing using Illumina Miseq. RESULTS: Mezavant provided the highest mucosal 5-ASA levels (geometric mean 2.39 ng/mg), followed by Asacol (1.60 ng/mg, 33% lower, P = 0.50) and Pentasa (0.57 ng/mg, 76% lower, P = 0.033). Mucosal 5-ASA concentration was not associated with NAT genotype, but serum 5-ASA concentration and NAT1 genotype was associated (P = 0.044). Mucosal 5-ASA concentration was positively associated with mucosal bacterial diversity (P = 0.0005) and bacterial composition. High mucosal 5-ASA concentration was related to reduced abundance of pathogenic bacteria such as Proteobacteria, and increased abundance of several favourable bacteria such as Faecalibacterium. CONCLUSIONS: Mucosal 5-ASA concentration is positively associated with bacterial diversity and a mucosal bacterial composition that are perceived favourable in UC. Mezavant yielded higher mucosal 5-ASA concentrations than Pentasa. 5-ASA may have beneficial effects on the mucosal microbiome, and high concentrations possibly amend dysbiosis in UC.