FDA’s December 2012 Guidance on IND and BA/BE Reporting – Part 2

FDA published in December 2012 a final guidance on IND and BA/BE reporting along with a brief Q&A accompanying it. See: www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM227351.pdf and www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM332846.pdf respectively.

Continuing with the review of the key points in FDA’s guidance.  This is the second of three postings.

Study Endpoints

If the endpoints are disease-related mortality or major morbidity or “safety endpoints”, the data should be evaluated by a DMC.  The protocol should pre-specify what will be reported as IND safety reports and what is not.  Cases that are specified in the protocol not to be reported should be reported as IND safety reports only if there is evidence suggesting a causal relationship between the drug & the AE.

Editorial Comment: No surprises here.

Anticipated SAEs & Protocol Specified Non-Reportable SAEs

FDA again stresses that they do not want SAEs sent in as IND safety reports that are anticipated to occur in a trial due to the patient population or disease studied.  Rather, they should be monitored at appropriate intervals and the number in each trial arm compared.  If there is an imbalance “suggesting there is a reasonable possibility that the drug caused the AE, this should be reported as an IND safety report.”

Other SAEs that the sponsor does not plan to report to FDA as IND safety reports should be noted in the protocol though FDA notes that it is not possible or desirable to list every AE that may occur in the study population.  Rather, a list of the more common SAEs, based on past experience, could be used which will make an overall safety analysis useful.  A description of how such cases will be monitored and submitted to FDA if a finding is made must be described.

Editorial Comment: Companies must now actively count and compare anticipated SAEs in each study arm and compare them and report if a significant imbalance for anticipated SAEs.  This may be “non-trivial” requiring the breaking of the blind by the statisticians and comparison with the other groups looking, presumably, at statistical comparisons and clinical meaningfulness.  It may be preferable to have the DMC handle these situations rather than risk the treatment groups being inadvertently revealed to the study team.  Clearly though, if a significant finding occurs, the study protocol, consent, IB etc. may need to be changed.

Non-Protocol Related SAEs

FDA makes some interesting comments about not reporting certain SAEs.  They say that not every single SAE not identified in the protocol must be an IND safety report and that medical judgment must be used.  In language that is a bit convoluted FDA then says:

“However, FDA will accept expedited reports for individual cases of unexpected SAEs that are not study endpoints and are not specified in the protocol as “unanticipated” to address concerns expressed by sponsors about not expeditiously reporting such cases.”

Editorial Comment: I read this as FDA saying, softly, we understand some companies will overreport cases that really don’t need to be 15 day safety reports, in order to cover themselves and that is OK…

Safety Surveillance for Ongoing Clinical Trials

FDA states that sponsors should have a systematic approach for safety surveillance which includes reviewing, evaluating, and managing accumulating safety data from the entire clinical trial database at appropriate intervals.  This may be within the company or external.

They remind sponsors that they must report expeditiously any findings from clinical, epidemiological, or pooled analysis of multiple studies (from any source – sponsor or not) or any findings from animal or in vitro testing that suggest a significant risk in humans exposed to the drug.

Sponsors must (where possible) look at the rates of SAEs and report any clinically important increase in the rate of an SAE compared to that listed in the protocol or IB always, of course, using medical judgment.

Editorial Comment: No surprises here.

Alternative Reporting Arrangements

Sponsor may request different reporting formats or frequencies if the FDA review division director agrees to this.  Conversely, FDA can also change the format or frequency or safety reporting (the company does not necessarily have to agree but does have to comply…).

Investigator Brochure (IB)

The FDA accepts a variety of formats for the IB but whatever format is used, the

IB should list those AEs observed with the drug for which a causal relationship is suspected or confirmed.  Class effects should also be listed.  AEs that are unlikely to have been caused by the drug should not be listed.

The IB should be updated in an ongoing basis either ASAP if important information needs to be conveyed or periodically (e.g. yearly) if the information is not critical.

FDA notes that there are many ways to do this and indicates that one way is to add a new serious unexpected suspected AR may be done as an addendum rather than reissuing the whole IB.

Editorial Comment:  The last paragraph is a bit of a surprise.  In the past some companies have added each 15 day report to the brochure routinely.  This involves sending the MedWatch/CIOMS I form to each investigator and instructing him or her to add this to the end of the brochure and to convey this to the IB if required and notifying the FDA of the IB update.  The reason companies do this is that now these cases (which are suspected of being causally related to the drug) now become listed and each subsequent case is no longer an expedited report.  This is a way to lower the number of expedited reports.  For some time this was frowned on as some form of “cheating” to avoid notifying FDA of more 15 day reports for these SAEs that are received.  But it is clear now that FDA accepts this methodology.  In a sense though nothing really changes as the FDA will now receive the MedWatch form as an IB update and not an expedited report.  So perhaps it is: plus ça change plus c’est la même chose (The more things change, the more they remain the same).

Unblinding (Unmasking)

FDA devotes much space to discuss the tricky issue of unblinding. (Note some companies, particularly makers of drugs for the eyes, do not like the words blinding and unblinding and prefer masking and unmasking, respectively).  FDA gives clear instructions:

• The blind should ordinarily be broken for IND safety reports submitted to FDA and all participating investigators.

• If the blind is broken and
  • The treatment is placebo this is not usually an expedited report.
  • The treatment is the test drug or active comparator, it must be an IND safety report

• For reporting aggregate analyses showing increased frequency in the drug treatment group, the analysis must include event rates in both the drug and placebo groups.
  • A systematic approach must be in place for evaluating accumulating safety data.  A DMC or an independent sponsor safety team can do this.

• If the sponsor believes the unblinding of AEs will compromise the integrity of the study, the sponsor should propose to the division director at FDA, in advance, an alternative reporting format or frequency to maintain the blind.

Editorial Comment:  The “surprise” here is that the 15 day unblinded cases sent to FDA must also be sent to all participating investigators.  I have seen companies that send the investigators blinded MedWatch/CIOMS I forms rather than unblinded ones.  This should not be done – everyone gets unblinded reports.

Investigator Reporting

Except for study endpoints, the investigator must immediately report to the sponsor all SAEs, whether he/she believes that they are drug related.  The investigator must report study endpoints that are SAEs in accordance with the protocol but not necessarily on the AE page of the case report form unless there is evidence suggesting a causal relationship.

Editorial Comment: No surprises here.

Causality

The FDA again notes that the investigator must include an assessment of causality for each SAE.  The FDA says the sponsor may decide how to capture this but gives an interesting example: a yes/no answer to the question is there a reasonable possibility that the drug caused the event?

Editorial Comment:  Many sponsors use a categorical approach with as many as five choices from a list such as this: related, probably related, possibly related, unlikely related, unrelated, not enough information).  There is no agreed upon standard so the choice is still the sponsor’s.

Non-Serious AEs (NSAEs)

NSAEs must be recorded and reported to the sponsor as specified in the protocol.  Investigator causality is not required though many sponsors request/require it.  For some studies NSAEs do not have to be collected (e.g. postmarketing outcome trials for a well established drug) if the FDA review division director approves.

Investigations of Marketed Drugs

As long as the sponsor maintains an IND for a drug, safety information from foreign and domestic studies, including non-IND studies, must be reported to the IND and to the NDA, if it meets the criteria for reporting.  If the sponsor receives a spontaneous report of an AE from U.S. or ex-US for a drug that is also under investigation, the report would not need to be reported to the IND because it is not a suspected AR observed in a study but it may be necessary to report it to the NDA.

Editorial Comment:  No surprises here but the perennial question of why a case must be reported to more than one part of FDA remains.  Presumably, FDA could communicate internally and share the data with those who need to know…

Reporting to IRBs

Investigators must promptly report “to the IRB…all unanticipated problems…including AEs that should be considered unanimated problems.

In general, a report that meets the criteria for reporting as an IND safety report would be an “unanticipated problem” and reported to the IRB by the investigator.

Some unanticipated problems may not be drug related (e.g. informed consent or privacy issues, certain adverse events that could not be caused by the investigational drug, such as events that occur prior to test article administration as a result of a washout period or due to a screening procedure) may or may not need to go to the IRB.  Such issues should be discussed with the FDA.

Editorial comment:  No surprises but not too useful…

Duration of Safety Reporting

Once a site has been officially closed out, the sponsor usually does not need to continue sending IND safety reports to the investigator since this information is normally used to protect patients in trials.  As there are no more patients in the trial, such reports need not be sent.  In unusual cases (e.g. delayed toxicity) such safety information may be reported to a site after it is closed out as the investigator may need to notify the concerned patients for follow up.

Editorial comment: No surprises.

The final part of this review will appear in part 3 in the next posting.