Study finds adverse reports inaccurate in cancer trials
November 1, 2009 4 minutes read
Study finds adverse reports inaccurate in cancer trials
Review of medical records shows inconsistencies
A study of serious adverse event reports (AERs) sent to an IRB from clinical trials of a breakthrough cancer drug revealed they were too often incomplete and inaccurate when compared to the original medical records from which they were taken.
The study, published earlier this year in the journal Clinical Cancer Research, suggests that incomplete reporting of serious adverse events could result in delays in identifying safety issues, which is especially problematic for cancer drugs on an accelerated approval path.1
But the advice from one of the study's authors for alleviating the problem may sound familiar: Stop sending a load of AERs to IRBs, and use specialized data centers for analysis. The result may be fewer, but more complete reports.
"Every IRB is basically getting inundated with a bunch of low-quality reports," says Charles L. Bennett, MD, PhD, MPP, co-director of the cancer control program at Northwestern University's Robert H. Lurie Comprehensive Cancer Center in Chicago." And then we try to ask each IRB to make a determination based on 40% completeness. That's just a bad way of going about it. We're not sure why the mandate is to have a huge quantity (of reports) at low quality. We should think about a system where we minimize the quantity and maximize the quality."
Comparing IRB reports to med charts
Bennett's group looked at serious adverse drug events reported to a single IRB from 14 clinical trials of the cancer drug imatinib, which received accelerated FDA approval in 2001 for the treatment of chronic myelogenous leukemia. The study looked at Phase I, II and III trials of the drug alone or in combination with other agents. Only 73 days elapsed from initial submission to the FDA to marketing approval, underlining the importance of identifying serious adverse drug events as quickly as possible.
In looking retrospectively at the serious AERs submitted to the IRB, Bennett's team also reviewed research charts and electronic patient medical records corresponding to each event.
The team created a standardized case report form that collected information about four types of events associated with the drug: infections, fractures, congestive heart failure/pulmonary effusion and pulmonary edema. For each event submitted to the IRB, the team used two versions of this case report form — one using data submitted to the IRB and the other based on the medical records of the same event.
A total of 115 separate imatinib-related serious adverse event reports were reviewed in this way. Researchers looked at how complete the information was — did it include all the information needed by the IRB to assess the seriousness of the event? When compared, the forms drawn from IRB reports were 40% complete, vs. 95% complete for forms based on the medical records.
In looking at the accuracy of the reports, in only 19% of the events did the IRB reports match the medical records completely.
But the review of the medical records revealed something even more striking: 22 potential serious adverse events that had not been reported to the IRB.
Bennett says it's important to note that the institution in question was a National Cancer Institute-designated comprehensive cancer center, one of only three sites where FDA licensing trials for imatinib were conducted.
"We picked a site where this drug was developed," Bennett says. "So if there was anybody who was going to know about the drug and what happens with it, it's this place. It's the best of the best and when we went to the best of the best, and we got what we got, it was really surprising."
Looking for stronger signals
Bennett suggests two possible reasons for the poor reporting of serious adverse events in the trials.
He says those reporting them are specialists in the disease being studied, but not necessarily in the types of events associated with the drug.
"Secondly, the adverse event reports are often filled out by staff people, research assistants," Bennett says. "They may not understand the clinical nuances of these adverse events and by not understanding the nuances, they may not do a good job of describing them."
He says the underreporting and lack of completeness can lead to a situation where early, subtle signals of a problem aren't strong enough to be noticed.
"You won't find anything with 40% completeness and 15% underreporting unless it's something really obvious."
In coming up with a system that better analyzes adverse events, Bennett suggests that NCI-designated centers could be set up to track trials, employing specialists who are formally trained in adverse events.
"We're setting up a program now where we do adverse event reporting at centers with medical toxicologists — they're going to do the work. They're not going to do every site in the country, but we have 38 big sites."
While there may be fewer reports, Bennett says, they will be much more complete and signals of problems will be clearer.
"At the end of the day, the goal is to make sure these clinical trials are providing appropriate and good information," Bennett says. "It's not important to get a huge quantity of these reports, it's most important to get the quality of these reports."
Reference
- Dorr DA, Burdon R, West DP et al. Quality of Reporting of Serious Adverse Drug Events to an Institutional Review Board: A Case Study with the Novel Cancer Agent Imatinib Mesylate. Clin Cancer Res 2009 Jun 1;15(11):3850-5.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content