The latest emergency response guidelines for clinical research

The safety of subjects’ lives comes first, the integrity of research data comes second, and compliance evidence is retained simultaneously. When there is a conflict between the three, priority is given to protecting the rights of subjects.

Last month I was watching the first human trial of CAR-T targeting BCMA at the Phase I clinical research center of a tertiary hospital in Guangzhou. The third subject developed grade 3 cytokine release syndrome (CRS) with high fever and sudden drop in blood pressure just 2 hours after the infusion. The old PI in charge of the project did not turn over half a foot of blood first as before. GCP standardizes the processing procedures. I directly took out my mobile phone and flipped through the CGT-specific disposal chapter of this new guideline. After checking the dosage and sampling requirements of tocilizumab, the subject was given the drug within 10 minutes, which was at least 15 minutes faster than following the process according to the 2020 version of the old guideline. In the end, the subject was saved and all data was not missed.

Don’t think that this guideline just adds a few pages of new treatments. This revision actually provides clear reference directions for several controversial points that the industry has been arguing about for many years. Let’s talk about the time limit for reporting emergencies. In the past, GCP specialists from many institutions “would rather over-report than under-report.” Whether the subjects were mildly nauseated after taking the medicine or really suffered from anaphylactic shock, they all went to the emergency reporting channel of CDE. Last year, a monoclonal antibody trial in a tertiary hospital in Shanghai produced a severe immediate allergic reaction. The reporting application was squeezed by hundreds of unrelated mild AE reports and took 2 hours to reach the reviewers, which almost delayed risk management and control. This new guideline directly clarifies the hierarchical reporting rules: only unexpected serious adverse events (SUSAR) determined to be related to the experimental drug need to be reported to CDE within 12 hours, and ordinary expected AEs can be submitted with the monthly research report. Of course, there are also conservative researchers who feel that this opening is a bit wide. They are afraid of being held responsible if they are missed in reporting. There are several centers around me that still require all SAEs to be reported within 24 hours. In fact, there is no problem. The guidelines only give minimum requirements. Institutions can adjust according to their own risk tolerance. There is no one-size-fits-all standard.

Oh, by the way, the most popular feature of this new guideline for multi-center research projects is the clarification of the compliance of digital emergency coordination systems. In the past, if SUSAR was discovered in a branch center in a cross-regional multi-center trial, the sponsor would have to send emails and call notifications to the PIs and ethics committees of dozens or even dozens of branch centers. It would take two or three days to fully cover it. If other centers enrolled subjects with the same high risk in the meantime, it would be easy for the same risk to be discovered. I participated in a national multi-center antidiabetic drug trial last month. A subject with low baseline blood sugar in the Shandong branch developed a hypoglycemic coma after taking the drug. The centralized emergency system used by the project team pushed event details and risk reminders to the mobile phones of all branch center PIs in just one minute. We adjusted the baseline blood sugar threshold in the inclusion criteria on the same day, and the same problem did not occur again in the future. There are also many veteran researchers who are not used to using this digital system and feel that notices signed on paper are more reliable. This is normal. The guide does not require that the system be changed, but only gives everyone a more efficient choice.

Speaking of no one-size-fits-all approach, the most controversial issue during this revision of the guidelines was the threshold for trial suspension. The conservative view is that as long as there is a case of grade 4 SAE related to the drug, the entire trial must be suspended to investigate risks. The radical view is that as long as there are clear intervention measures, the subjects can recover, and the risk incidence does not exceed the preset value, enrollment can continue. The new guidelines do not take any side, but give a decision-making framework: whether to suspend the trial, you must look at three dimensions at the same time: whether the SAE is unexpected, whether there are effective intervention methods, and whether the incidence rate exceeds the risk threshold preset in the plan. Only when all three conditions are met, a comprehensive suspension needs to be considered. Otherwise, you can just adjust the entry criteria or the dosage. The most extreme example I have seen is an anti-tumor drug trial last year. A case of grade 4 thrombocytopenia occurred. The conservative ethics committee directly requested that the trial be suspended. In the end, it was found that the subject had a basic blood system disease, which was not related to the drug. After the evaluation was carried out according to the framework of the new guidelines, the trial was resumed after only 3 days of suspension. If it had been based on the old rules, it might have been suspended for several months.

To be honest, no matter how the guidelines are updated, they are essentially for clinical researchers. In an urgent situation, how can you have time to read every clause? The first thing to take care of must be the subjects lying on the bed. This is also the core of all clinical research specifications that has never changed from the birth of GCP to the present. Oh, by the way, if you are still using the 2020 version of the old guideline, change it quickly. Last month, a sponsor was directly given a warning letter by CDE because it failed to report the neurotoxicity SUSAR of the CGT trial in time as required by the old guideline. The project was stopped for almost half a year, and it suffered a huge loss.