Becker's Hospital Review

75 CIO / HEALTH IT UPenn Researchers Want to Launch the US' 1st CRISPR Trial in Humans: 6 Things to Know By Jessica Kim Cohen A team of researchers from Philadel- phia-based University of Pennsylva- nia expect to receive FDA clearance to conduct the first CRISPR trial treating hu- man patients in the U.S., according to e Wall Street Journal. However, due to a less stringent regulatory process, physicians in China have been run- ning human trials with the gene-editing tool since as early as 2015. ere are nine human CRISPR trials from China listed in a U.S. Na- tional Library of Medicine database. Here are six things to know about human tri- als using CRISPR in the U.S. and China. 1. CRISPR, which stands for Clustered Regular- ly Interspaced Short Palindromic Repeats, is a gene-editing technology that enables scientists to edit an organism's DNA. Many scientists con- sider the CRISPR-Cas9 system — which creates modified RNA segments that bind to the CRIS- PR-associated protein 9 enzyme — to be one of the most precise and least expensive gene-edit- ing techniques currently in use. 2. Scientists in the U.S. helped develop CRIS- PR-Cas9, which researches have successfully used to edit genomes in animals, such as mice suffering from hereditary deafness, and even in human cells in laboratory environments. However, physicians in the U.S. have not been cleared by regulatory agencies to use the tech- nology in human trials. 3. e researchers at UPenn have been work- ing toward their human CRISPR trial — which would edit DNA from cancer patients to improve their ability to fight the disease — for almost two years to meet various federal requirements to ensure patient safety. e re- searchers completed the study's ethics review in late 2017 and are awaiting final FDA clear- ance, which they expect as early as January. 4. Dr. Wu Shixiu, an oncologist and president of Hangzhou Cancer Hospital in China, has been using CRISPR-Cas9 to treat cancer pa- tients by deleting a gene that may inhibit the immune system's ability to fight the disease since March. Dr. Wu did not need approval from national regulators for the trial, and his hospital review board reportedly took one af- ternoon to OK the trial. "China shouldn't have been the first one to do it," Dr. Wu told e Wall Street Journal. "But there are fewer restrictions." 5. While China's regulatory process has few- er roadblocks, many researchers in the U.S. emphasized ethical considerations regarding gene editing, given CRISPR makes irrevers- ible changes in the human genome. While Dr. Wu agrees there are risks to the technology, he noted the cancer patients he sees are facing terminal diseases. "If we don't try, we will nev- er know," he told e Wall Street Journal. 6. e Wall Street Journal noted the U.S. re- searchers it spoke with didn't suggest Ameri- ca relax its regulatory requirements. Instead, they tended to advocate for an "international consensus" on ethical issues related to these emerging technologies that change human DNA, particularly as researchers are still studying potential unintended consequences. "How do we make sure everyone is under the same tent?" said Jeffrey Kahn, PhD, director of the Berman Institute of Bioethics at Balti- more-based Johns Hopkins University. "We need to be talking to each other internationally." n Google Predicts Medical Outcomes With 46B Data Points, Artificial Intelligence: 5 Things to Know By Jessica Kim Cohen A team of Google researchers developed an artificial intelligence system that predicts medical outcomes like mortality and readmissions based on patient data held in EHRs, Quartz reports. The team's results were published in a research paper sub- mitted to the scientific preprint website arXiv.org Jan. 24. The research paper has not been peer-reviewed. Here are five things to know about the research project. 1. To construct a predictive model, researchers typically ex- tract select variables from standardized data. However, to avoid what Google researchers referred to as a "labor-in- tensive process," they attempted to use deep learning, an advanced machine learning technique that doesn't require standardized data. 2. The researchers developed a deep learning approach to analyze a patient's raw EHR record. The deep learning system evaluated data from thousands of patients to deter- mine words and events associated with certain outcomes, while also identifying which data could be ignored. 3. To validate the deep learning approach, researchers used de-identified EHR data from 216,221 adult patients who were hospitalized for 24-plus hours at UC San Fran- cisco Medical Center and University of Chicago Medicine. This volume of EHR data encompassed 46,864,534,945 data points, including clinical notes. 4. The deep learning models achieved high accuracy for predicting in-hospital mortality, 30-day unplanned read- missions, prolonged length of stay and all of a patient's fi- nal discharge diagnoses, according to the research paper. 5. The study authors concluded their deep learning ap- proach outperformed existing "state-of-the-art traditional predictive models." "We believe that this approach can be used to create ac- curate and scalable predictions for a variety of clinical sce- narios, complete with explanations that directly highlight evidence in the patient's chart," the study authors wrote. n

• Cover