No abstract

Background As there is limited data on the sustainability of desensitization of multifood-oral immunotherapy (multifood-OIT), we conducted a multisite multifood-OIT study to compare the efficacy of successful desensitization with sustained dosing vs discontinued dosing after multifood-OIT. Methods We enrolled 70 participants, aged 5–22 years with multiple food allergies confirmed by double-blind placebo-controlled food challenges (DBPCFCs). In the open-label phase of the study, all participants received omalizumab (weeks 1–16) and multi-OIT (2–5 allergens; weeks 8–30) and eligible participants (on maintenance dose of each allergen by weeks 28–29) were randomized 1:1:1 to 1 g, 300 mg, or 0 mg arms (blinded, weeks 30–36) and then tested by food challenge at week 36. Success was defined as passing 2 g food challenge to at least 2 foods in week 36. Findings Most participants were able to reach a dose of 2 g or higher of each of 2, 3, 4, and 5 food allergens (as applicable to the participant's food allergens in OIT) in week 36 food challenges. Using an intent-to-treat analysis, we did not find evidence that a 300 mg dose was effectively different than a 1 g dose in maintaining desensitization, and both together were more effective than OIT discontinuation (0 mg dose) (85% vs 55%, P = 0.03). Fifty-five percent of the intent-to-treat participants and 69% of per protocol participants randomized to the 0 mg arm showed no objective reactivity after 6 weeks of discontinuation. Cross-desensitization was found between cashew/pistachio and walnut/pecan when only one of the foods was part of OIT. No statistically significant safety differences were found between the three arms. Interpretation These results suggest that sustained desensitization after omalizumab-facilitated multi-OIT best occurs through continued maintenance OIT dosing of either 300 mg or 1 g of each food allergen as opposed to discontinuation of multi-OIT. Funding Sean N. Parker Center for Allergy and Asthma Research at , , AADCRC U19AI104209. Trial Registration Number ClinicalTrials.gov number, NCT02626611 .

Asthma affects nearly 300 million people worldwide. The majority respond to inhaled corticosteroid treatment with or without beta-adrenergic agonists. However, a subset of 5 to 10% with severe asthma do not respond optimally to these medications. Different phenotypes of asthma may explain why current therapies show limited benefits in subgroups of patients. Interleukin-13 is implicated as a central regulator in IgE synthesis, mucus hypersecretion, airway hyperresponsiveness, and fibrosis. Promising research suggests that the interleukin-13 pathway may be an important target in the treatment of the different asthma phenotypes.

BACKGROUND AND AIMS: Endoscopic disease activity scoring in ulcerative colitis (UC) is useful in clinical practice but done infrequently. It is required in clinical trials, where it is expensive and slow because human central readers are needed. A machine learning algorithm automating the process could elevate clinical care and facilitate clinical research. Prior work using single-institution databases and endoscopic still images has been promising. METHODS: Seven hundred and ninety-five full-length endoscopy videos were prospectively collected from a phase 2 trial of mirikizumab with 249 patients from 14 countries, totaling 19.5 million image frames. Expert central readers assigned each full-length endoscopy videos 1 endoscopic Mayo score (eMS) and 1 Ulcerative Colitis Endoscopic Index of Severity (UCEIS) score. Initially, video data were cleaned and abnormality features extracted using convolutional neural networks. Subsequently, a recurrent neural network was trained on the features to predict eMS and UCEIS from individual full-length endoscopy videos. RESULTS: The primary metric to assess the performance of the recurrent neural network model was quadratic weighted kappa (QWK) comparing the agreement of the machine-read endoscopy score with the human central reader score. QWK progressively penalizes disagreements that exceed 1 level. The model's agreement metric was excellent, with a QWK of 0.844 (95% confidence interval, 0.787-0.901) for eMS and 0.855 (95% confidence interval, 0.80-0.91) for UCEIS. CONCLUSIONS: We found that a deep learning algorithm can be trained to predict levels of UC severity from full-length endoscopy videos. Our data set was prospectively collected in a multinational clinical trial, videos rather than still images were used, UCEIS and eMS were reported, and machine learning algorithm performance metrics met or exceeded those previously published for UC severity scores.