Treatment of severe Coronavirus Disease 2019 (COVID-19) is challenging. We performed a phase 2 trial to assess the efficacy and safety of human umbilical cord-mesenchymal stem cells (UC-MSCs) to treat severe COVID-19 patients with lung damage, based on our phase 1 data. In this randomized, double-blind, and placebo-controlled trial, we recruited 101 severe COVID-19 patients with lung damage. They were randomly assigned at a 2:1 ratio to receive either UC-MSCs (4 × 107 cells per infusion) or placebo on day 0, 3, and 6. The primary endpoint was an altered proportion of whole lung lesion volumes from baseline to day 28. Other imaging outcomes, 6-minute walk test (6-MWT), maximum vital capacity, diffusing capacity, and adverse events were recorded and analyzed. In all, 100 COVID-19 patients were finally received either UC-MSCs (n = 65) or placebo (n = 35). UC-MSCs administration exerted numerical improvement in whole lung lesion volume from baseline to day 28 compared with the placebo (the median difference was −13.31%, 95% CI −29.14%, 2.13%, P = 0.080). UC-MSCs significantly reduced the proportions of solid component lesion volume compared with the placebo (median difference: −15.45%; 95% CI −30.82%, −0.39%; P = 0.043). The 6-MWT showed an increased distance in patients treated with UC-MSCs (difference: 27.00 m; 95% CI 0.00, 57.00; P = 0.057). The incidence of adverse events was similar in the two groups. These results suggest that UC-MSCs treatment is a safe and potentially effective therapeutic approach for COVID-19 patients with lung damage. A phase 3 trial is required to evaluate effects on reducing mortality and preventing long-term pulmonary disability. (Funded by The National Key R&D Program of China and others. ClinicalTrials.gov number, NCT04288102.

Coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is spreading rapidly around the world, resulting in a massive death toll. Lung infection or pneumonia is the common complication of COVID-19, and imaging techniques, especially computed tomography (CT), have played an important role in diagnoses and treatment assessment of the disease. Herein, we review the use of imaging characteristics and computing models that have been applied for the management of COVID-19. CT, positron emission tomography -CT (PET/CT), lung ultrasound, and magnetic resonance imaging (MRI) have been used for detection, treatment, and follow-up. The quantitative analysis of imaging data using artificial intelligence (AI) is also explored. Our findings indicate that typical imaging characteristics and their changes can play an important role in the detection and management of COVID-19. In addition, AI or other quantitative image analysis methods are urgently needed to maximize the value of imaging in the management of COVID-19.

In this paper, we introduce XGLUE, a new benchmark dataset that can be used to train large-scale cross-lingual pre-trained models using multilingual and bilingual corpora and evaluate their performance across a diverse set of cross-lingual tasks. Comparing to GLUE (Wang et al., 2019), which is labeled in English for natural language understanding tasks only, XGLUE has two main advantages: (1) it provides 11 diversified tasks that cover both natural language understanding and generation scenarios; (2) for each task, it provides labeled data in multiple languages. We extend a recent cross-lingual pre-trained model Unicoder to cover both understanding and generation tasks, which is evaluated on XGLUE as a strong baseline. We also evaluate the base versions (12-layer) of Multilingual BERT, XLM and XLM-R for comparison. 1