Two recent issues have appeared documenting late reporting of clinical trial information (including safety data) to the FDA and other federal agencies. This is a timely reminder to all those in the business of reporting safety information to FDA and other agencies to ensure that this is being done in a complete and timely manner.
National Cancer Institute
In October of this year, the National Cancer Institute (NCI) reported that some of their lymphoma researchers failed to report serious adverse events (deaths) in one of its studies to the FDA in a timely manner. The information was reported in the press, including the Wall Street Journal.
Dr. Francis Collins, the NIH director, indicated that there were two deaths from aspergillus in 2015 in a trial of Ibrutinib in CNS lymphoma being done at the NIH hospital. The reporting to FDA did not occur until May 2016, according to the article. IRB reporting was also late. Full details were not available but Dr. Collins noted that a “severe (sic) adverse event needs to be reported” within a matter of days to FDA. This refers to the reporting of IND safety events as 7 or 15 day reports. Presumably, Dr. Collins meant serious, not severe AEs.
The article notes that when the first death occurred it was not considered “especially noteworthy. But when a second death took place in 2015…the conclusion was that the drug cocktail — including steroids…and ibrutinib — led to a likely extreme immune suppression that contributed to the deaths.”
Dr. Collins indicated he “plans to ask all the institutes [of NIH] to see if there have been cases of severe adverse events in clinical trials that weren’t reported to the FDA or other authorities in a timely fashion.”
A similar story appeared in the Washington Post with further details, in which it was noted: “Officials said that after the second death, the principal investigator… expressed some concerns to the part of the NCI that was sponsoring the trial and the institutional review board overseeing the study. In addition, he ordered CT scans in other patients in the trial to try to catch fungal infections early, before they became deadly. As concerns mounted, researchers stopped enrolling new patients in April this year. But…did not file an official “unanticipated problem” report — discussing the suspected relationship between the infections and the treatment — until May this year.”
The Washington Post further noted that “after being notified this spring about the connection between the treatment and the deadly infections, [the FDA] conducted a review that concluded there also were several other adverse events that weren’t promptly reported.” In addition, [the investigator], who has been suspended from clinical research until he undergoes additional training, did not respond to email and telephone messages requesting comment. To ensure researchers know and comply with reporting requirements, Collins…said, training sessions for other researchers also will be conducted, as will audits of various trials.”
This situation is, unfortunately, not unknown in the world of clinical research. The usual problems occur in clinical trials sponsored by pharmaceutical companies but, as we see here, it can also occur in studies done by government bodies. In the typical scenario, an investigator fails to report SAEs and even deaths to the sponsor as required by regulation and by the study contract signed with the company as well as forms submitted to FDA promising to comply with the requirements.
What is not totally clear is whether the reports were sent by the investigator to the sponsor (or whoever was overseeing the clinical trial safety as it is not clear in the news reports who should inform FDA in this trial) and whether it was judged by the investigator and the sponsor to be possibly related to the study drug. Presumably, aspergillus infection was not an expected (labeled) event. So it is not clear where the breakdown occurred. It also seems that the investigator took action by doing CT scans of other patients to see if they had infections. It is not clear whether this action was discussed with the sponsor, FDA, the IRBs and whether the patients themselves were informed by means of a revised informed consent. It is not clear if there was a Data Monitoring Board/Committee (DMC) external to the trial also overseeing patient safety and, if yes, whether they were informed.
In this scenario in pharmaceutical sponsored trials, it would not be uncommon for the sponsor to immediately send a monitor or auditor to the site, as well as possibly to other sites to look for unreported SAEs and to find the root cause of the problem. The late reports would immediately be sent to FDA, the investigators and the IRB as required. The information would be reviewed by the company and action would be taken rapidly including perhaps a phone discussion FDA, discussion with the DMC and an urgent plan put in place. This might indeed include suspension of further enrollment and/or treatment of ongoing patients as well as scans looking for Aspergillus (as was done here). Additional training would surely be done also if there was a failure to report SAEs.
It is also quite possible that FDA would do a surprise, unannounced inspection (audit) of the site and possibly the sponsor to ensure that the safety procedures in place were adequate and being followed.
Unfortunately, I was not able to find out any other information beyond what is in the two news reports. I could find nothing on the FDA, NCI, NIH and ClinicalTrials.gov websites. Perhaps it is there, but I could not find it.
As Dr. Collins noted, this is a very serious issue and one hopes that a good root cause analysis will be done to see if the issue is isolated or systemic. It is also hoped that a formal corrective action/preventive action (CAPA) program has been put in place. One presumes that there is a Quality Management System in place to handle this. One would like to see the results of the investigation and what steps are being put in place to prevent future issues. It is also hoped that the results of all of this will be reported to the public and easily available on the websites.
The second report on problems in reporting is from Oxford University’s Evidence-Based Medicine Data Lab in the UK. In October 2016 they updated their ongoing report entitled “Who’s not sharing their trial results.” In this program they regularly review all interventional trials (except for phase 0/1 and those with a formal request to delay results) registered on ClinicalTrials.gov. This report covers January 2006 to August 2016. The full data is also available online.
The full reference is: Powell-Smith A and Goldacre B. The TrialsTracker: Automated ongoing monitoring of failure to share clinical trial results by all major companies and research institutions [version 1; referees: awaiting peer review]. F1000Research 2016, 5:2629.
Clinicaltrials.gov is a website run by the US NIH as a registry for global clinical trials and which includes information on the protocol, the sites, as well as safety and efficacy data. In a nutshell, sponsors must report the required data within 21 days of the first participant’s enrollment in the trial and the results of the trial must be submitted within 12 months after the completion date.
The Oxford University researchers found that 45% of the 25, 927 trials are missing study results. This has been consistent since 2006. See graph.
They also name names, including the top non-compliers. There are 290 institutions on the list including almost all of the major academic, governmental and pharmaceutical company research institutions. The full listing is available at the URL above and the complete dataset is also available (though not very user-friendly). Only one institution seems to have 100% compliance.
This is not news and, unfortunately, it is old news. It has been clear for quite some time that compliance is not good. The FDA and some congressmen have commented on this. Articles in the pharma literature, such as this one, have appeared. Yet nothing much seems to have changed.
Note, by the way, that the NCI is number three on this list with an overall 35% non-compliance rate on clinicaltrials.gov. In fairness, however, drilling down on the data shows that they have markedly improved over the last several years. The bulk of the lateness was in 2006 and 2007. Nonetheless, it seems there are some reporting issues at the NCI. Another irony is that one branch of the NIH (NCI) is not reporting data to another branch of the NIH (clinicaltrials.gov).
It seems that unless FDA or Congress puts in place penalties for non-compliance, the problem will not be resolved.
For those in the pharmacovigilance field, the lesson is, as always, that drug safety must be done well and done in a timely manner. This is most visible and easily tracked by companies and FDA for expedited IND and NDA reports and PSURs and DSURs. But other requirements in place for patient safety must also be followed assiduously.