Explore the vast range of titles available.

The advent of anti-TNF agents has dramatically changed the treatment algorithms for IBD in the last 15 years, but primarily and more importantly secondary loss of response is often observed. Fortunately, new treatment options have been actively explored and some have already entered our clinical practice. In the class of anti-cytokine agents, the anti-IL12/IL23 monoclonal antibodies (mAbs) have entered clinical practice with the anti-p40 mAb ustekinumab in Crohn’s disease (CD). Also, more selective anti-IL23 agents (anti-p19) have shown efficacy and are being further developed, in contrast to agents inhibiting IL-17 downstream which have failed in clinical trials despite their clear efficacy in psoriasis (Verstockt et al. in Expert Opin Biol Ther 17(1):31–47, 2017; Verstockt et al. in Expert Opin Drug Saf 16(7):809–821, 2017). Following up on the efficacy of the anti-adhesion molecule vedolizumab, etrolizumab (anti-beta-7 integrin) and PF-00547659, an anti-MadCam mAb, are being developed (Lobaton et al. in Aliment Pharmacol Ther 39(6):579–594, 2014). Oral anti-trafficking agents, such as ozanimod, targeting the S1P receptor responsible for the efflux of T-cells from the lymph nodes, have also shown efficacy in patients with ulcerative colitis (UC) (Sandborn et al. in N Engl J Med 374(18):1754–1762, 2016). Oral agents inhibiting cell signaling have been explored successfully in IBD. Tofacitinib, a non-selective oral Janus kinase (JAK) inhibitor, is effective in patients with UC and several other more or less selective Jak1, 2 and 3 inhibitors are being developed for the treatment of CD and UC (Sandborn et al. in N Engl J Med 376(18):1723–1736, 2017; Vermeire et al. in Lancet 389(10066):266–275, 2017; De Vries et al. in J Crohns Colitis 11(7):885–93, 2017). Finally, despite initial disappointing results with systemic administration of mesenchymal stem cells, Alofisel, adipose tissue derived, allogeneic mesenchymal stem cells, locally injected in perianal fistula tracts, induce long-lasting beneficial effects and the drug has been approved in Europe (Panes et al. in Gastroenterology, 2017). In summary, the quest for new treatment options in IBD is very active and justified by the high medical need and unresolved problems patients are facing.

Endometrial disorders represent a major gynaecological burden. Current research models fail to recapitulate the nature and heterogeneity of these diseases, thereby hampering scientific and clinical progress. Here we developed long-term expandable organoids from a broad spectrum of endometrial pathologies. Organoids from endometriosis show disease-associated traits and cancer-linked mutations. Endometrial cancer-derived organoids accurately capture cancer subtypes, replicate the mutational landscape of the tumours and display patient-specific drug responses. Organoids were also established from precancerous pathologies encompassing endometrial hyperplasia and Lynch syndrome, and inherited gene mutations were maintained. Endometrial disease organoids reproduced the original lesion when transplanted in vivo. In summary, we developed multiple organoid models that capture endometrial disease diversity and will provide powerful research models and drug screening and discovery tools. Boretto et al. demonstrate that organoids derived from patients with various types of endometrial pathologies can model disease traits and diversity, and can be used as a drug-screening tool.

Ulcerative colitis (UC) is a chronic inflammatory bowel disease of unknown aetiology affecting the colon and rectum. Multiple factors, such as genetic background, environmental and luminal factors, and mucosal immune dysregulation, have been suggested to contribute to UC pathogenesis. UC has evolved into a global burden given its high incidence in developed countries and the substantial increase in incidence in developing countries. An improved understanding of the mechanisms underlying UC has led to the emergence of new treatments. Since the early 2000s, anti-tumour necrosis factor (TNF) treatment has significantly improved treatment outcomes. Advances in medical treatments have enabled a paradigm shift in treatment goals from symptomatic relief to endoscopic and histological healing to achieve better long-term outcomes and, consequently, diagnostic modalities have also been improved to monitor disease activity more tightly. Despite these improvements in patient care, a substantial proportion of patients, for example, those who are refractory to medical treatment or those who develop colitis-associated colorectal dysplasia or cancer, still require restorative proctocolectomy. The development of novel drugs and improvement of the treatment strategy by implementing personalized medicine are warranted to achieve optimal disease control. However, delineating the aetiology of UC is necessary to ultimately achieve disease cure. Ulcerative colitis is a chronic inflammatory bowel disease affecting the rectum and the colon. This Primer describes the epidemiology, the current understanding of ulcerative colitis pathophysiology and its diagnosis. In addition, Hibi and colleagues highlight the current and emerging therapeutic strategies and discuss the most pressing questions in the field.

ObjectivesTo develop and validate a proof-of-concept convolutional neural network (CNN)–based deep learning system (DLS) that classifies common hepatic lesions on multi-phasic MRI.MethodsA custom CNN was engineered by iteratively optimizing the network architecture and training cases, finally consisting of three convolutional layers with associated rectified linear units, two maximum pooling layers, and two fully connected layers. Four hundred ninety-four hepatic lesions with typical imaging features from six categories were utilized, divided into training (n = 434) and test (n = 60) sets. Established augmentation techniques were used to generate 43,400 training samples. An Adam optimizer was used for training. Monte Carlo cross-validation was performed. After model engineering was finalized, classification accuracy for the final CNN was compared with two board-certified radiologists on an identical unseen test set.ResultsThe DLS demonstrated a 92% accuracy, a 92% sensitivity (Sn), and a 98% specificity (Sp). Test set performance in a single run of random unseen cases showed an average 90% Sn and 98% Sp. The average Sn/Sp on these same cases for radiologists was 82.5%/96.5%. Results showed a 90% Sn for classifying hepatocellular carcinoma (HCC) compared to 60%/70% for radiologists. For HCC classification, the true positive and false positive rates were 93.5% and 1.6%, respectively, with a receiver operating characteristic area under the curve of 0.992. Computation time per lesion was 5.6 ms.ConclusionThis preliminary deep learning study demonstrated feasibility for classifying lesions with typical imaging features from six common hepatic lesion types, motivating future studies with larger multi-institutional datasets and more complex imaging appearances.Key Points• Deep learning demonstrates high performance in the classification of liver lesions on volumetric multi-phasic MRI, showing potential as an eventual decision-support tool for radiologists.• Demonstrating a classification runtime of a few milliseconds per lesion, a deep learning system could be incorporated into the clinical workflow in a time-efficient manner.

Complex perianal fistulas in Crohn's disease are challenging to treat. Allogeneic, expanded, adipose-derived stem cells (Cx601) are a promising new therapeutic approach. We aimed to assess the safety and efficacy of Cx601 for treatment-refractory complex perianal fistulas in patients with Crohn's disease. We did this randomised, double-blind, parallel-group, placebo-controlled study at 49 hospitals in seven European countries and Israel from July 6, 2012, to July 27, 2015. Adult patients (≥18 years) with Crohn's disease and treatment-refractory, draining complex perianal fistulas were randomly assigned (1:1) using a pre-established randomisation list to a single intralesional injection of 120 million Cx601 cells or 24 mL saline solution (placebo), with stratification according to concomitant baseline treatment. Treatment was administered by an unmasked surgeon, with a masked gastroenterologist and radiologist assessing the therapeutic effect. The primary endpoint was combined remission at week 24 (ie, clinical assessment of closure of all treated external openings that were draining at baseline, and absence of collections >2 cm of the treated perianal fistulas confirmed by masked central MRI). Efficacy was assessed in the intention-to-treat (ITT) and modified ITT populations; safety was assessed in the safety population. This study is registered with ClinicalTrials.gov, number NCT01541579. 212 patients were randomly assigned: 107 to Cx601 and 105 to placebo. A significantly greater proportion of patients treated with Cx601 versus placebo achieved combined remission in the ITT (53 of 107 [50%] vs 36 of 105 [34%]; difference 15·2%, 97·5% CI 0·2–30·3; p=0·024) and modified ITT populations (53 of 103 [51%] vs 36 of 101 [36%]; 15·8%, 0·5–31·2; p=0·021). 18 (17%) of 103 patients in the Cx601 group versus 30 (29%) of 103 in the placebo group experienced treatment-related adverse events, the most common of which were anal abscess (six in the Cx601 group vs nine in the placebo group) and proctalgia (five vs nine). Cx601 is an effective and safe treatment for complex perianal fistulas in patients with Crohn's disease who did not respond to conventional or biological treatments, or both. TiGenix.

Model-informed precision dosing (MIPD) software tools are used to optimize dosage regimens in individual patients, aiming to achieve drug exposure targets associated with desirable clinical outcomes. Over the last few decades, numerous MIPD software tools have been developed. However, they have still not been widely integrated into clinical practice. This study focuses on identifying the requirements for and evaluating the performance of the currently available MIPD software tools. First, a total of 22 experts in the field of precision dosing completed a web survey to assess the importance (from 0; do not agree at all, to 10; completely agree) of 103 pre-established software tool criteria organized in eight categories: user-friendliness and utilization, user support, computational aspects, population models, quality and validation, output generation, privacy and data security, and cost. Category mean ± pooled standard deviation importance scores ranged from 7.2 ± 2.1 (user-friendliness and utilization) to 8.5 ± 1.8 (privacy and data security). The relative importance score of each criterion within a category was used as a weighting factor in the subsequent evaluation of the software tools. Ten software tools were identified through literature and internet searches: four software tools were provided by companies (DoseMeRx, InsightRX Nova, MwPharm++, and PrecisePK) and six were provided by non-company owners (AutoKinetics, BestDose, ID-ODS, NextDose, TDMx, and Tucuxi). All software tools performed well in all categories, although there were differences in terms of in-built software features, user interface design, the number of drug modules and populations, user support, quality control, and cost. Therefore, the choice for a certain software tool should be made based on these differences and personal preferences. However, there are still improvements to be made in terms of electronic health record integration, standardization of software and model validation strategies, and prospective evidence for the software tools' clinical and cost benefits.

Intestinal barrier defects are common in patients with inflammatory bowel disease (IBD). To identify which components could underlie these changes, we performed an in-depth analysis of epithelial barrier genes in IBD.MethodsA set of 128 intestinal barrier genes was selected. Polygenic risk scores were generated based on selected barrier gene variants that were associated with Crohn's disease (CD) or ulcerative colitis (UC) in our study. Gene expression was analyzed using microarray and quantitative reverse transcription polymerase chain reaction. Influence of barrier gene variants on expression was studied by cis-expression quantitative trait loci mapping and comparing patients with low- and high-risk scores.ResultsBarrier risk scores were significantly higher in patients with IBD than controls. At single-gene level, the associated barrier single-nucleotide polymorphisms were most significantly enriched in PTGER4 for CD and HNF4A for UC. As a group, the regulating proteins were most enriched for CD and UC. Expression analysis showed that many epithelial barrier genes were significantly dysregulated in active CD and UC, with overrepresentation of mucus layer genes. In uninflamed CD ileum and IBD colon, most barrier gene levels restored to normal, except for MUC1 and MUC4 that remained persistently increased compared with controls. Expression levels did not depend on cis-regulatory variants nor combined genetic risk.ConclusionsWe found genetic and transcriptomic dysregulations of key epithelial barrier genes and components in IBD. Of these, we believe that mucus genes, in particular MUC1 and MUC4, play an essential role in the pathogenesis of IBD and could represent interesting targets for treatment.

The interleukin-23 pathway is implicated genetically and biologically in the pathogenesis of Crohn's disease. We aimed to assess the efficacy and safety of risankizumab (BI 655066, Boehringer Ingelheim, Ingelheim, Germany), a humanised monoclonal antibody targeting the p19 subunit of interleukin-23, in patients with moderately-to-severely active Crohn's disease. In this randomised, double-blind, placebo-controlled phase 2 study, we enrolled patients at 36 referral sites in North America, Europe, and southeast Asia. Eligible patients were aged 18–75 years, with a diagnosis of Crohn's disease for at least 3 months, assessed as moderate-to-severe Crohn's disease at screening, defined as a Crohn's Disease Activity Index (CDAI) of 220–450, with mucosal ulcers in the ileum or colon, or both, and a Crohn's Disease Endoscopic Index of Severity (CDEIS) of at least 7 (≥4 for patients with isolated ileitis) on ileocolonoscopy scored by a masked central reader. Patients were randomised 1:1:1 using an interactive response system to a double-blind investigational product, and stratified by previous exposure to TNF antagonists (yes vs no). Patients received intravenous 200 mg risankizumab, 600 mg risankizumab, or placebo, at weeks 0, 4, and 8. The primary outcome was clinical remission (CDAI <150) at week 12 (intention-to-treat population). Safety was assessed in patients who received at least one dose of study drug. This study is registered with ClinicalTrials.gov, number NCT02031276. Between March, 2014, and September, 2015, 213 patients were screened, and 121 patients randomised. At baseline, 113 patients (93%) had been previously treated with at least one tumour necrosis factor (TNF) antagonist (which had failed in 96 [79%]). At week 12, 25 (31%) of 82 risankizumab patients (pooled 41 patients in 200 mg and 41 patients in 600 mg arms) had clinical remission versus six (15%) of 39 placebo patients (difference vs placebo 15·0%, 95% CI 0·1 to 30·1; p=0·0489). Ten (24%) of 41 patients who received 200 mg risankizumab had clinical remission (9·0%, −8·3 to 26·2; p=0·31) and 15 (37%) of 41 who received the 600 mg dose (20·9%, 2·6 to 39·2; p=0·0252). 95 (79%) patients had adverse events (32 in the placebo group, 32 randomised to 200 mg risankizumab, 31 randomised to 600 mg risankizumab); 18 had severe adverse events (nine, six, three); 12 discontinued (six, five, one); 24 had serious adverse events (12, nine, three). The most common adverse event was nausea and most common serious adverse event was worsening of underlying Crohn's disease. No deaths occurred. In this short-term study, risankizumab was more effective than placebo for inducing clinical remission in patients with active Crohn's disease. Therefore, selective blockade of interleukin-23 via inhibition of p19 might be a viable therapeutic approach in Crohn's disease. Boehringer Ingelheim.

Fecal calprotectin is a marker of inflammation in inflammatory bowel disease (IBD). Since mucosal healing has become a goal of treatment in IBD we examined how reliably calprotectin levels reflect mucosal disease activity.MethodsIn all, 126 IBD patients and 32 irritable bowel syndrome (IBS) patients needing colonoscopy delivered a sample of feces prior to the start of bowel cleansing. Besides collection of symptom scores and blood tests, experienced endoscopists recorded the Simple Endoscopic Score for Crohn's Disease (SES-CD) and the Crohn's Disease Endoscopic Index of Severity (CDEIS) in Crohn's disease (CD) patients and the Mayo endoscopic score in ulcerative colitis (UC) patients. Stool samples were shipped for central calprotectin PhiCal Assay (enzyme-linked immunosorbent assay [ELISA]). Correlation analysis was done with Pearson statistics.ResultsThe median (interquartile range [IQR]) fecal calprotectin levels were 175 (44–938) μg/g in CD, 465 (61–1128) μg/g in UC, and 54 (16–139) μg/g in IBS. Correlations were significant with endoscopic disease scores in both CD and in UC. Using ROC statistics, a cutoff value of 250 μg/g indicated the presence of large ulcers with a sensitivity of 60.4% and a specificity of 79.5% (positive predictive value [PPV] 78.4%, negative predictive value [NPV] 62.0%) in CD. Levels ≤250 μg/g predicted endoscopic remission (CDEIS ≤3) with 94.1% sensitivity and 62.2% specificity (PPV 48.5%, NPV 96.6%). In UC, a fecal calprotectin >250 μg/g gave a sensitivity of 71.0% and a specificity of 100.0% (PPV 100.0%, NPV 47.1%) for active mucosal disease activity (Mayo >0). Calprotectin levels significantly correlated with symptom scores in UC (r = 0.561, P < 0.001), but not in CD.ConclusionsFecal calprotectin levels correlate significantly with endoscopic disease activity in IBD. The test appears useful in clinical practice for assessment of endoscopic activity and remission.

ObjectivesTo develop a proof-of-concept “interpretable” deep learning prototype that justifies aspects of its predictions from a pre-trained hepatic lesion classifier.MethodsA convolutional neural network (CNN) was engineered and trained to classify six hepatic tumor entities using 494 lesions on multi-phasic MRI, described in Part 1. A subset of each lesion class was labeled with up to four key imaging features per lesion. A post hoc algorithm inferred the presence of these features in a test set of 60 lesions by analyzing activation patterns of the pre-trained CNN model. Feature maps were generated that highlight regions in the original image that correspond to particular features. Additionally, relevance scores were assigned to each identified feature, denoting the relative contribution of a feature to the predicted lesion classification.ResultsThe interpretable deep learning system achieved 76.5% positive predictive value and 82.9% sensitivity in identifying the correct radiological features present in each test lesion. The model misclassified 12% of lesions. Incorrect features were found more often in misclassified lesions than correctly identified lesions (60.4% vs. 85.6%). Feature maps were consistent with original image voxels contributing to each imaging feature. Feature relevance scores tended to reflect the most prominent imaging criteria for each class.ConclusionsThis interpretable deep learning system demonstrates proof of principle for illuminating portions of a pre-trained deep neural network’s decision-making, by analyzing inner layers and automatically describing features contributing to predictions.Key Points• An interpretable deep learning system prototype can explain aspects of its decision-making by identifying relevant imaging features and showing where these features are found on an image, facilitating clinical translation.• By providing feedback on the importance of various radiological features in performing differential diagnosis, interpretable deep learning systems have the potential to interface with standardized reporting systems such as LI-RADS, validating ancillary features and improving clinical practicality.• An interpretable deep learning system could potentially add quantitative data to radiologic reports and serve radiologists with evidence-based decision support.