Good Clinical Practice: A Question & Answer Reference Guide - May 2016 - 375

Drug/Study Safety and Safety Reporting
administrative trials (Grant AM, Altman DG, Babiker AB, Campbell MK, et al. DAMOCLES study group, "Issues
in data monitoring and interim analysis of trials," Health Technology and Assessment, 2005;9(7):1-238).
In the IND safety reporting final regulation issued in September 2010, the FDA states that a DSMB is not
required and is not necessary for most studies, particularly those evaluating symptomatic treatments. Generally,
DSMBs are associated with large, randomized multisite trials that are designed to evaluate treatments intended to
improve survival or reduce the risk of major morbidity. In such cases, the independent DSMB would be expected
to monitor serious events that are study endpoints and also may assess the rate of other known consequences of
the underlying disease or other events that are common in the study population. FDA's DSMB guidance also notes
another potential use for a DSMB. Some sponsors have used DSMBs to monitor the overall event rates as the safety
database accumulates in ongoing studies (see FDA DSMB guidance). A DSMB could periodically analyze and
evaluate the aggregated, unblinded events in the entire IND safety database to determine if the drug is the suspected
cause. During these analyses, investigators and study participants would remain blinded.
9.48 Q. When a DSMB is used, are its records considered part of the sponsor's records? Can the FDA inspect,
and do IRBs have the authority to require the submission and review of, DSMB meeting minutes and
other related records?
A. While the FDA has the authority to inspect any organization participating in a clinical trial conducted under an
IND, the agency does not currently inspect DSMBs because it does not have regulations requiring their routine
use. An IRB's reach does not extend to inspecting sponsors or, by extension, DSMBs that may be advising
them. Further, DSMBs are not monitored by clinical research associates and are not typically audited. They are
autonomous, independent entities that generally are not subject to external review. Typically, DSMB records are
considered confidential, and their decisions are disclosed only to sponsors and possibly to a committee of study
investigators. Since a DSMB is set up and chartered (and paid) by the sponsor, they can establish what documents
and information will be retained and who will have access to this information.
Although the FDA has never inspected a DSMB or its records, it could theoretically seek access to DSMB records
as part of a follow-up to a sponsor inspection. The agency might do so, for example, if it suspected that a DSMB
made an important safety-related recommendation that a sponsor did not report to the FDA, or if it suspected that
a DSMB was being pressured to under-report safety-related concerns to the sponsor. Agency officials concede,
however, that the lack of regulations regarding DSMB inspections and the lack of an agency DSMB inspectional
program make the FDA's authority in this area less than definitive. The agency would have clear authority to inspect
a DSMB if the sponsor formally transferred certain regulatory obligations to the board, however.
9.49 Q. Are hospitalizations that occur more than 30 days after the end of the study period reportable to the
FDA? In some conditions, should AEs that occur after a subject has gone off study be captured on case
report forms?
A. The study period during which all AEs and SAEs must be reported generally begins after informed consent
is obtained and the initiation of study treatment (or initiation of any study procedures) and ends after a protocolspecified post-treatment follow-up period. In the absence of specific delayed toxicity concerns or safety hypotheses,
a post-treatment follow-up period of at least five half-lives of the drug under investigation is often recommended.
With respect to identifying deaths that may be relevant to a product's safety review, the FDA's clinical safety
review guidance mentions that a four-week period of follow-up after a patient leaves a study or otherwise
discontinues study drug is reasonable for drugs with prompt action and relatively short elimination half-lives.
For drugs with particularly long elimination half-lives or drug classes with a recognized potential to cause lateoccurring effects, deaths occurring over longer timeframes following drug discontinuation should be evaluated.
The appropriate duration of post-treatment safety follow-up will depend on the characteristics of the investigational
product and can be discussed with applicable regulatory authorities (e.g., during end-of-phase-2 meetings).
After the protocol-specified post-treatment follow-up period, investigators are generally instructed to report only
SAEs that they suspect could be attributed to prior study treatment. Some protocols require post-study survival
follow-up or assessments of other post-study clinical outcomes as part of efficacy analyses. These data would
generally not be considered AEs/SAEs.

375

Table of Contents for the Digital Edition of Good Clinical Practice: A Question & Answer Reference Guide - May 2016