Latest revision of clinical research emergency response guidelines

In July 2024, the Center for Drug Evaluation (CDE) of the State Food and Drug Administration and the National Center for Infectious Diseases issued the latest revised version of the "Clinical Research Emergency Response Guidelines". The core adjustments focus on the three major directions of "dynamic adaptation of hierarchical response thresholds", "priority boundary refinement of subject rights and interests" and "standardization of cross-regional multi-center collaborative processing". For the first time, the emergency response exemption rules for special clinical research scenarios such as rare diseases and advanced tumors have been clarified. It has now been officially implemented in drug clinical trial institutions (GCP institutions) nationwide.

I was doing project monitoring at a phase I clinical center of a tertiary hospital in Shanghai last week. I happened to encounter the institution’s quality control teacher leading a CRC (clinical research coordinator) training on updating the new guidelines. There was a review of an adverse event in a subject that was just handled last month on the screen. In the old 2020 version of the guidelines, this case would most likely be considered a "process non-compliance", but in the new version, it has become a benchmark case.

To be honest, the old version of the guideline has been in use for 4 years, and there have been a lot of complaints from the clinical side. Especially PIs (principal investigators) who are conducting oncology and rare disease trials are troubled every time they encounter adverse events with vague boundaries: according to the old version, as long as a grade 3 adverse reaction is reached, the trial drug must be stopped immediately and an ethics report must be submitted within 24 hours, but many patients with advanced tumors are enrolled. All standard treatments have been tried before, and experimental drugs are the only way to survive. Stopping the drugs rashly may make the progress faster. Previously, a PI in Guangzhou was given a rectification notice by the regulatory authorities because he added a liver-protective drug to the patient and did not stop the drug immediately. He was wronged for a long time.

When the public opinion consultation meeting was held for this revision, the two sides argued quite fiercely. Experts on the regulatory side feel that process standards must be tightened, otherwise it will be easy for someone to take advantage of the loopholes and use the rights of subjects as a cover to advance experiments in violation of regulations. In the end, no one will be held responsible if something goes wrong. ； Experts on the clinical side cited a bunch of cases of rare disease trials, saying that one-size-fits-all rules really do not treat subjects as human beings. In the end, both sides took a step back, and there is now the "individualized exemption clause": as long as the PI can produce clear evidence of benefit assessment and sign to confirm that the benefits to the subject of continued medication are far greater than the risks, the medication can be temporarily suspended and submitted to the ethics and monitoring department for review. As long as the follow-up materials are correct, it will not be considered a compliance issue.

For example, in my Phase II trial of ALK-positive advanced lung cancer, there was a 52-year-old female subject who was resistant to chemotherapy and immunotherapy before enrollment, and brain metastases appeared. After taking the experimental drug for 8 months, the tumor shrank by 90%. Last month, a reexamination suddenly revealed a grade 3 elevated transaminase. For the old version, we would definitely stop her medication immediately. For the new version, the PI assessed her liver function indicators and the risk of tumor progression, talked with her family for half an hour, and finally signed the application for exemption. She was given dual liver-protective drugs, and she continued to test the drugs. She checked her liver function twice a week. After 10 days, the index dropped to level 1. When she came for a review yesterday, she told us that she could finally go back to pick up her grandson from school.

Of course, not everyone is satisfied with the new version. I had dinner with the CRC of a small SMO (Clinical Research Site Management Organization) a while ago, and they complained that the cross-center collaboration requirements in the new version are too troublesome: as long as one center in a multi-center trial has a serious emergency, the ethics, researchers, and monitors of all centers must be synchronized within 24 hours. They have seven or eight projects on hand at the same time, and the manpower is not enough. Sometimes when the ethics committee in a remote area is off work and no one answers the phone, it makes my hair fall out. There are also some institutions where the ethics committee process is slow, and no feedback is given for half a month after receiving the exemption application. The PI does not dare to make a decision casually, but is stuck in a dilemma.

I have worked in clinical supervision for 6 years, and I have seen CROs (Contract Research Organizations) that conceal adverse events in order to speed up the trial progress, and I have also seen PIs that argue with ethics in order to obtain medication opportunities for subjects. To be honest, what touches me most about this revision is not how many new clauses have been added, but the addition of the sentence "the rights of subjects take precedence over trial progress and compliance procedures" in the general principles. In the past, everyone's first reaction when encountering an emergency was "will they be fined?"

The new version has just been implemented for three months now, and various institutions are still adjusting, and many practical details have not been completely smoothed out. I heard that CDE will hold an online Q&A session next month, and will definitely make up practical details based on everyone's feedback. To put it bluntly, there are never perfect rules for clinical research. They are all adjusted little by little according to the actual problems encountered. As long as the direction is right, it doesn't matter if it is slower.