动物实验 z-bo 2020-09-26 352

　　Many animal studies are very irregular, and the more rigorous such studies, the more applicable the results are to human biology.

　　A paper being studied by Ulrich Dirnagl, director of the Stroke Research Center at Charité Medical School in Berlin, Germany, explains how the new drug can treat rodent brains after stroke. The author used 20 mice as experimental subjects, 10 of which received this treatment. However, it is difficult to understand that only 7 treated mice appeared in the analysis result graph.

　　Dirnagl recalled: "I wrote to the editor and said that I could not judge this paper. I need to know where the other three mice went." Six months later, the editor replied: it is. He learned from the author of the paper. , These three mice died of a stroke. The author just omitted the three mice in the paper. However, further analysis of the stroke drugs revealed important information about the death of these three mice. In addition to reducing the effect, this treatment also causes brain damage.

　　Dirnagl said: "This is not a hoax." Dirnagl's research often uses mice as experimental subjects. He explained that there are many reasons why the tested animals are not included in the research results, but this is a phenomenon recognized by the academic community. "Sometimes it's just a whim, and people exclude experimental animals without reporting." Dirnagl thinks this is a bad habit.

　　For several years, researchers, pharmaceutical companies, drug regulators, and even the public have hardly realized the contribution of animals in the development of drugs for the treatment of human diseases. Pay attention to whether the symptoms of mice with various diseases can be accurately reflected in people with this disease. But Dirnagl and some people say there is another equally serious problem. They think that many animal studies are very irregular. The more rigorous such research, the better the results obtained in human biology.

　　Of course, it is difficult to achieve "one size fits all". Animal research covers a wide range of biological fields, from single molecule activity in healthy organs to the possible side effects of new drugs being tested in humans. Many animal research enthusiasts also accept a truly responsible attitude and consider ways to organize experiments wisely and carefully.

　　Even animal studies that have a significant impact on human drug research follow far fewer standards than human clinical trials (such as studies reviewed by Dirnagl). Looking at clinical trials, you will see a long list of "include" and "exclude" criteria to manage who can become a volunteer. If you have high blood pressure or cancer and are taking certain medications, you are not eligible to participate in volunteer activities.

　　However, animal research rarely follows these rules. Out of ethical and cost considerations, researchers strive to use as few animals as possible. This means minimizing the sample size. The operation of the experiment was normal. In the words of Dirnagl, "The way we conduct animal research seems to have gone back to the Stone Age."

　　The reason for the deviation

　　Ten years ago, Malcolm Macleod, a neuroscientist at the University of Edinburgh in the UK, was looking for new treatments for stroke. McLeod and his colleagues conducted animal experiments with 603 drugs, of which 374 may help treat brain diseases. Among them, 97 drugs have been tested in humans, but only one was found to be effective.

　　Surprised by the huge difference between the results of the animal and human experiments, MacLeod began to wonder which part of the error occurred. He speculates that human treatment trials may not have been conducted under appropriate conditions, such as a low dose or a long time after a stroke. Another speculation is that human testing has been completed correctly, but the selected animal is not a suitable subject for studying human stroke. The third possibility is that the researchers did not test the drug correctly at the beginning of the animal experiment.

　　In this regard, McLeod conducted a more detailed investigation, and the results shocked him. Only 36% of animal studies included random assignment of stroke medication or placebo to animals. In addition, McLeod said, these non-randomized studies severely overestimated the effectiveness of these drugs.

　　Macleod then turned to study neurological diseases: Alzheimer's disease, multiple sclerosis, Huntington's disease, Parkinson's syndrome, etc. When conducting animal studies to find possible treatments, the actual situation is worse than studying stroke cases. Since MacLeod's tutor and tutor follow the same method, many authors do not seem to realize the importance of lack of rigor. Lisa Bello, an expert on health policy at the University of California, San Francisco, said: “I was trained to be an animal researcher. The random choice they understand is to put your hand in a cage and grab one that happens to rush to you. Animals around. In other words, choosing animals is not a random method. "Some animals may be startled or chewed, or they may curl up and fall asleep. It is at this point that the deviation appears.

　　Status review

　　As a biophysicist, Shanghai Silberberg, program director of the National Institute of Neurological and Stroke (NINDS) in Bethesda, Maryland, received training in Israel and the United States three years ago. He found that NINDS gave the green light to projects that were not scientific facts. He decided to investigate the matter.

　　Silverberg decided to start with animal research. To some extent, he needs to process data similar to MacLeod contacts. The truth surprised him: animal studies are very biased. Part of the INDS budget ranges from animal research to human testing. Silberberg worries that the suffering of these "poor patients" should be avoided.

　　In the summer of 2012, after failing to negotiate with the National Institutes of Health (NIH), Silberberg and like-minded researchers held a seminar in Washington. All participants are magazine editors. Silberberg believes that journal editors play an important role in improving the standards of animal research. He said: “In the beginning, everyone blamed each other and took responsibility. The editor was responsible, the review was responsible, and the funding agency was responsible. On the last day of the meeting, everyone agreed. When it was reached, it was everyone’s responsibility.”

　　Since then, NIH has strengthened the review of all projects funded by animal research. Many organizations are beginning to experiment with new review methods. Lawrence Tabak, deputy dean of the National Institutes of Health, said: "Some basic principles, such as the basis of experimental design and the basis of statistics, may be forgotten."

　　Another attempt of

　　IH was carried out at the National Institute of Environmental Health Sciences in North Carolina. Here, toxicologists are looking for ways to classify animal studies. This classification allows regulatory agencies to make correct recommendations for specific chemicals. The US Environmental Protection Agency is also considering methods for evaluating animal research data.

　　academic journals also joined this initiative. In April, "Nature" published a list of authors and reviewers, requesting that the scientific method in life science papers should include more details. In June, "Science Translation Medicine" announced a similar initiative.

　　However, some people think that these requirements are unnecessary. Joseph Bass, an obesity and diabetes expert at Northwestern University in Chicago, Illinois, said: "I don't think the entire scientific community will cover up the facts and systemic biases discovered." The problem is that he does not believe that data is the main cause of bias. On the contrary, he believes that the reasons for deviation in different fields and different types of experiments are different. For example, in the field of bathing research, metabolism is affected by temperature, and animals happen to be very sensitive to temperature changes. If genetic manipulation has side effects on animals, the experimental results will also be biased. He believes that it is impractical to apply uniform requirements to all animal studies.